So it’s been just over 6 months since I retired. I thought I’d write up a 6 month check in and report in on how things are going. First of all - and point for my first point - I 100% recognized how lucky and blessed I am to be able to retire “early”. And that is the first thing I have noticed - I have Retirement Guilt - I get this occasionally interacting with people. Sometimes it’s something they directly say : “Why is it you are retired and am I still working?” (My Mom) and others are due to me comparing my situation with theirs and feeling guilty that they are working and I am not. This is part of life always - “Never compare yourself to others - or you will become vain or bitter” - and the song Funny the Way It Is - exemplifies the irony of life :

Funny the way it is, if you think about it

One kid walks ten miles to school, another’s dropping out

Funny the way it is, not right or wrong

On a soldier’s last breath his baby’s being born

To be clear it’s most often from myself - but there is often a subtle undercurrent of surprise when people ask what I am doing. And to further clarify I don’t have regret about retiring; work never defined me : I worked hard and had my opportunity to make impacts from individual to systemic - but I always had a full life outside of work. And I was lucky to both love my job and to have a job that I didn’t hate 100% of the time, or even a majority of the time.

So to one of the most often asked questions : Am I bored? - No - not at all - I always have plenty to do. I have always had a bunch of hobbies (outdoors : trail running, biking, hiking, backpacking, mountaineering and photography and videography) and add to that a long list of things to take care of and do and I easily fill up my day. My normal activities can now take priority for scheduling, for example : rather than get up at 5:30 am to go on a run before a work meeting - I can now get up at 6:30 am and take a bit more time getting out the door and go on my run.

What else have I been doing with my time? One of the main themes here is that I have expanded some of my hobbies. Including doing too much activity and injuring myself. For example I have really gotten into shooting 35mm film - I have bought a number of old used cameras, and shot a lot of film. For example, I took a personal project to photograph Western States in June all on film : click and manually crank the wind of the shot and then shoot again - I did that for 14 rolls, you can see the results here. And then after paying the bill for development and seeing how long it takes to get film back - I decided I’d try developing film myself - you can see my first scans here. It is quite a process - but I’ll cover that in a separate post. Another hobby whose nature is large by its very nature : I also did another 50K in April - traversing Zion National Park with Seth before he turned 50.

The second theme is doing more things myself - things I used to pay others to do - that now that I have time I can do myself, learn new things and have a sense of satisfaction in completing them.

I started mowing my own lawn; I bought an electric leaf blower and a weed whacker. With Rod’s help, I also planted a bunch plants, spread some mulch, got the sprinklers going, and replaced the sprinkler controller with a Rachio - that gave me app/wifi control over timing and zones. I also wasn’t afraid of using local teenager help to do some of the labor, especially after injuring myself. ;)

I also bought a bike stand, and some tools and am attempting to teach myself how to repair/tweak my bike vs. having a shop do everything. Refreshed the toobeless goop in my tires, adjusted the derailer and minor things like that.

I also built a sauna, which involved me doing extensive research on how to build a true wet Finnish sauna with all the important details like insulation, venting, bench size and height etc etc. I built it with my general contractor buddy Steve, where he did all the real specialized work : measuring, cutting, nailing etc. And I paid for everything and was the hired laborer : “Carry those boards down to the back porch.” - “Stain all those cedar boards on the exterior and interior” - or provided specialized sauna research : “What gauge of wire do you buy and install for 200 degree heat?” ( 8 AWG Electrical Silicone Copper Core High Temperature Resistant Wire) and what kind of exhaust fan do you buy? We started May 4th and finished July 16th - another topic that I need to do a more extensive write up about. Thankfully I have really enjoyed using the sauna 3-4 times a week since finishing it.

We have traveled a bit - mostly locally - road trip to Arizona to see the grandkids, and then through Joshua Tree, Black Canyon, Grand Canyon, Zions and Logan. A trip to San Francisco for a good friend and co-workers memorial. Another trip to Arizona. And then a trip as noted to Western States.

I have also been able to attend more closely to helping Stac with her medical appointments. It’s a slow process working your way through the medical system from GP to specialist to specialist, but we are making progress bit by bit.

Film teaches you about photography in a very different way. I think it helps immensely to have had the digital tutelage about ISO/aperture/shutter speed and the instant feedback to help you improve. But film makes you slow down - and really make sure you have things right. As Joe Grant said about film in 2018 on iRunFar.

I needed to learn how to look, to observe, to pay attention to a scene, emphasizing the importance of composition. … I needed to practice… [to] … spend time learning the craft, underlining the value of the process.

So I started shooting a lot more film and I decided I wanted to try to photograph Western States but entirely on film. I built a pretty detailed plan (with the consultation of Gemini) on where I’d shoot, what camera and lenses I’d use and what film stock. My camera shelf is an expression of the N+1 philosophy - but I wanted to keep things as simple as possible. So I opted to only take 4 film cameras and 3 lenses.

Leica M6 with the Voigtlander Nokton 50mm f/1.1 lens.

Olympus OM-2S with a OM Zuiko 28mm f/2.8 and a OM Zuiko 300mm f/4.5

All three of these lenses were manual focus - so it would not be as easy as “point and shoot” or spray and pray even as I’d have to manually wind the film after each shot.

The other two camera were both point and shoot Fuji Klasse - the Standard (S) with a 35mm lens and the Wide (W) with a 28mm lens.

My rough plan and execution went as follows :

Friday - hang out at the Western States Expo for bib pickup and Pre-Race Meeting and try to catch folks to take their portraits. I had emailed a bunch of athlete managers, brand sponsor contacts and tagged a bunch of folks on Instagram - but response was limited or a flat out “no - we are not doing any more media requests”. I planned to use the Leica with a roll of Porta 400 and the Klasse 35 with Kodak Tri-X 400. Things are pretty chaotic and I wasn’t able to get that many portraits - but I was able to corner a few folks who all graciously said yes. Killian was fast and illusive, could never catch him, even though Nick Danielson tried, we just missed each other. Jim, I got lucky by happening to be hanging out with the Jurek’s who schemed with Jim to pull off this classic post. After the race briefing - I picked up my livestream equipment for my duties at Quarry Road the next day.

Saturday - I woke at 3 am and got ready and made my way to the track. I was shooting the Olympus OM-2S with the 28mm f/2.8 and the Leica M6 50mm f/1.1, both had Ilford 3200 ISO film loaded. Almost every single one of these 60 photos was a bust. Light meter kept reading the headlamps and the spot lights and everything was under exposed and dark. I wouldn’t know that though until weeks later when I got the film back from the lab.

From there I drove to Devils Thumb and setup a tripod with the Olympus OM-2S using the 300mm f/4.5 - the trick was try to catch the runners coming up into Devils Thumb and hit focus. I started by focusing on a spot (using a stump) that I could time the shot with as they runner arrived and then pull back past them and then try to snap the photo and they ran into focus. Each shot had to be manually wound, pull focus all while keeping my eye to the view finder and trying to time the shot with when they were in focus. After the fact I was happy with getting quite a few shots, but missing a bunch more as well. After the runner was past my tri-pod - I’d have my Fuji Klasse-W in my hand and run back around the bushes to the aid station and snap a few shots with the point and shoot.

Around 1 - I left Devils Thumb and drove the dusty and bumpy road back to Forest Hill. Somewhere in the dust my tire hit something and as I pulled onto the paved Forest Hill road my tire went flat. I pulled over to the side and turned on my blinkers and got out to see about changing a flat. I crawled under the truck (a Dodge Ram 1500 rental) and there was a full size tire “screwed” into some fastening device under the truck. But I couldn’t find a jack or anyway to get the tire. Nothing in the back seat, nothing under the truck, nothing in the bed. So I stood there on the side of the road, there was no cell service - no way to call anyone. A lady stopped in her van with 3 kids and a “trunk” full of luggage - she was crewing her husband who was running and she was heading for Forest Hill. She saw my livestream vest and pulled over and offered me a ride. I grabbed my camera bags, film and the huge box full of livestream gear and climbed into her front seat with it all on my lap.

At Forest Hill - I arrived after the lead men had ran through but the lead women were still to come. I was able to catch a few photos with the Fuji Klasse-W but I was majorly distracted trying to find a tow truck to fix the flat or tow the truck so that I could get to Quarry road by 4pm. After calling 5 tow trucks, being told yes and then an hour later being told no. I texted Billy and asked him if Tim Tollefson (who I had seen at Devils Thumb) would want to drive me out to Quarry Road. Tim texted me- we met and he graciously agreed to drive me and the gear out to Quarry Road. We had the gate code and we drove the old gravel road that was rough in a couple of places but his trusty Subaru got us through. Hal Koener and Scott Jurek run the aid station - and this year Jenny Jurek whipped up the Quarry Pitt theme, which was a playful nod to the Circle Pit and added a self effacing bit of levity. The lead men were running hot - ahead of course record pace and we barely got the live stream camera set up before Vincent came running through.

The top men and women all came through in blurried pace of succession. The aid station swarmed each one and got them ready to get on with the last 10 miles. After seeing the top 10 women - Tim and I left to drive back to the High School. As we walked back toward the track - Puppi was curled up in a plastic chair beside the road with his crew - looking very tired. Tim crouched down to congratulate him on an epic historic 2nd place sub-14 finish. Back at the track - I called AAA for a tow truck and watched a few of the 10 ten women finishers and wandered around the field where the runners took a well needed break. I got to see Jeff Browning cross the line just after 11pm, breaking the 50 year old masters finishing record.

Finally at 11:30 pm AAA called and said they had a truck coming. Just past midnight the Tow Truck called saying he was coming into Auburn from Colfax. BUT he said he’d have to meet me at Forest Hill grocery store - but he couldn’t give me a ride - as he had someone else with him that he needed to go tow after me. He was worried if something went wrong with not being able to change the tire - I’d just be stranded 40 miles down Forest Hill road. I saw Scott Sowle on the field and asked him if he’d be willing to drive me out there; he graciously agreed. We arrived at 1 am and rendezvoused with the tow truck, and it took us another 40 minutes to do the long drive out to where my truck was flat and because I left the flashers on was now completely dead. We unlocked the car with the physical key hidden in the fob, jumped the truck and then the driver found the jack and the extension rods to lower the tire under the passenger front seat. (!?) By 2 am he had the tire changed and I was on my way. It was an hour drive back to Auburn and there was only a couch in the AirBnb. I was dead tired - I’d been awake 23 hours and I didn’t feel like trying wake up my brother-in-law at 3:30 am - so I called the airport - changed my flight from Monday to Sunday at 6:15 am that day and flew back home.

It was nerve wracking having the TSA hand check my 14 rolls of film instead of putting them through the scanner. And even more frustrating when one roll triggered a security check, which they firmly informed me and reaffirmed with a supervisor meant they had to put that one roll through the scanner (it still turned out!).

I mailed the film off the theFINDLab on Monday and waited patiently over the 4th of July weekend to get the scans back. Not knowing what I’d get. I shot 14 rolls of film at States, 2 of those were the partials - finished on the way there and after I got returned. There were 689 scans of Western States; in the editing process - I marked 242 of them as worthy of review. 184 were worth a 2nd look. I narrowed that down to 56 - 4 star and finally picked 38 to feature on my website - you can see all 132 on Smugmug.

I have been helping a few of my classmates prepare for my 40th High School reunion this fall. (Sept 2026) I have created some registrations forms, done some digital sleuthing to try to located classmates and scanned in a 7th grade Jr. High year book and our Senior year book photos. So many faces are familiar and so many I can’t even recall. High School is this thing that none of us choose to go into per se but is part of the growing up process that your just sort of propelled into. Depending on when you move into an area, you go to elementary school with a small class, then you go to a middle school and then finally are all put into an even bigger class in High School. My class was around 250. Among all of those classmates and across the four grades - we all gather each day of the week for school : classes, teachers, lunches, PE and sports and extra curricular activities.

Everyone had their groups - either from friendships, or based on a common class or activity. I was in band and marching band for the first 3 years. I had friends from Jr. High that did not persist into High School (Billy Bjorkland) and others that did (David Jones, Blake Casselman). In the 80s there were the jocks, the cheerleaders, the band nerds, the smart nerds, the cowboys, the rockers, the tough guys etc etc. We all intermixed in between classes and at lunch and at sporting events and dances - we were all just trying to stumble our way through our high school teenage years. For sure some knew where they were headed and had everything together. Others maybe were just “Working for the Weekend” and some (like me) were kinda lost and unsure for much of it. Looking back on the faculty photos - 40 years later I realize that while at the time we saw them as the adults to be tolerated, ignored or adored they were all just trying to live their own middle aged lives - to have a career - to make a living.

None of us are defined by High School - we are all who we are today based on the choices we have made since; the events that have unfolded in our lives, who we’ve loved and where we’ve lived. And yet 40 years ago we spent every school day for year(s) together - and while we didn’t know everyone at the same level of detail - it’s always good to get back together and to remember those days and to try to find the familiar in the faces 40 years later.

I was in a room with a group of scientists - and there was a small machine that looked like an African kalimba thumb piano - and it had produced a series of anomalies that had caused a change in the time continuum. Alterations had been initiated and changes had occurred.

“Well I have a problem with this” - said Miles pointing to the bulging sharp protrusion in his stomach that visibly pushed through his t-shirt.

“He is going to give birth to a pterodactyl and there is nothing that we can do about it” - said one of the faceless scientists. “But there will be people who will try to stop you.”

We went to a travel motel that had the rooms arranged in one half of a rectangle; the parking lot was a light pea gravel. It was dark and the beings from the past were starting their attacks. They were from the 1400s and had flails and swords. We fought them off valiantly - at the windows with their removable screens and the hollow core doors. All the while checking on Miles and his coming pterodactyl. The fighting went on for much of the night - with waves of marauders, chasing, evading and conquering.

We eventually ended up at a Catholic church - that had an open nave, and a balcony that encircled it on 3 sides with stairways at either end. The final fighter had been sent from the past. He was a large man, a knight with short dark hair and a very large sword with which he fought and delivered death. I saw Wright Noel, who was dressed in a western cowboy hat, and I asked him “Are you packing?” - he said no but then he disappeared.

The knight began making his way up the stairs - wielding his sword, slashing and destroying. In the top hallway he made his way towards me and suddenly Wright rushed up the stairs carrying a large Colt revolver, when he reached the balcony he fired and the knights head exploded in a bright flange of the darkest red - frozen in a volcanic state of eruption.

I stumbled out of my subconscious state to half awake and thought - “That was a very weird dream - I need to remember that.”

This is not about software or work and it will be a long read; hopefully it’s an interesting read - in the event you want to do the Zions Traverse or understand how I undertake attempts at long distance events on trails however mostly it’s an opportunity for me to reflect. For sure its a record of why in the last half of my career I didn’t feel afraid of anything at work.

Seth wanted to do a 50k before he turned 50. Like all feats of endurance it’s part of a curiosity about whether you can. Added to that is the opportunity to explore beautiful countryside along the way with good friends.

Originally we were gonna tackle Rim to Rim to Rim on the Grand Canyon, but the fire and the North Rim closure shut that down. Spring was our timeframe and so I turned to other south western destinations where we could take advantage of good weather - warm but hopefully not too hot.

I have had my eye on the Zion Traverse of Zion’s National Park (ZNP) for a while and mapped out the route. Originally Seth had wanted to do 50 miles but readjusted for a 50k and the Zion Traverse checked that box at around 60k. The traverse starts at Kolob Canyon - the far western side of Zions about 20 miles south of Cedar City and makes its way south westerly through canyons and over open country to the White Rim and down past the infamous Angels Landing to the park’s main road.

Logistics

The logistics of a point to point traverse always require coordination. Someone to drop you off and some way to have a car or someone to pick you up at the end. The driving distance between the start and the finish is only an hour - but if your aiming to have a car at the finish you have to drive all the way into ZNP first and then back track. Bureaucracy and regulations make this tricky. The main road into ZNP closes to cars every spring and fall and transportation is limited to shuttle busses. The last shuttle leaves the park at 7:15 pm - so if you don’t make that - you have another 5-6 miles out of the park into town. There is a lodge in the park, that is only .6 miles from the end of the trail but rooms are hard to come by, though I got lucky to catch a cancellation for the night our journey would complete. Staying at the Lodge does give you a code that lets you open the gate and drive to the Lodge. HOWEVER - if you happen to need to park a car at the Lodge the day prior to your arrival - they refuse to give you a parking pass, because you can of course only get a parking pass for the day you are checking in. We needed to leave a car there on the day before. The Lodge refused to budge and said we should ask the rangers for a yellow permit that they give to climbers who are driving in - parking and climbing. However the rangers said because we were not actually entering the park from the eastern side and were starting at the western side we didn’t qualify for a permit. They wouldn’t issue one to just drive in and leave a car.

Bureaucracy and regulations will never bend to the law of common sense. Between a rock and a hard place - Since we wouldn’t have a car when we arrived I drove into the Lodge to drop off bags so we’d have fresh clothes to change into. Only to be informed by the Lodge that they didn’t check bags as a matter of policy. Though a nice lady at front advocated for us and did in fact check out bags. As I filled out the bag tags - I apologized because I had 3 bags - she replied “Oh its ok, since we don’t take bags it doesn’t matter to me how many you check.” :)

I drove back out to meet Seth at the visitor center where he’d acquired our valid wilderness permit to enter at Lees Pass and traverse the park and we set out for the hour drive to the trail head. We were both blown away by how amazing the towering red rock formations were in Kolob Canyon - and its only a 3 minute drive off the freeway - well worth the stop! Our plan was to camp at the trailhead and then start out early Tuesday morning. Of course the park doesn’t allow you to sleep in your car or camp at the trail head - regulations again! We managed to find a flat spot up the hill out of sight and laid out our pads and bags to cowboy camp. After watching a memorable sunset fade across the red rocks that only glow even more red in the waning light - we went to bed pretty early. I fell asleep immediately but woke up around midnight for the old man pee. The moon was more than 3/4s full and was so bright. I did not sleep well - I kept waking up - not because I was worried or anxious or even that I was cold but just kept waking up. I’d read some (The Rush), played a game or listened to a podcast until I fell back asleep only to wake back up some time later. Finally at 4:30 am I was starting to feel the cold and I had to pee again - so at 4:45 am I yelled at Seth that I was getting up.

We packed up and wandered back down to the car to make some breakfast (greek yogurt, granola and raspberries) and get our packs together.

What I packed

My favorite long distance (non-overnight) running pack was the Ultimate Direction Mountain Vest V4, sadly that pack broke when I was testing things out Sunday (the vest sides have four “ropes” that tie to a back support and those ropes broke - I was/am sad). So I ended up using a smaller UD Mountain Vest of the same pack that lacked the side pouches and the back kangaroo pouch. Years ago I’d had Shon Hancock sew some side pockets on. The pack is about 12 liters with a main pouch and the typical shoulder straps for bottles and gels. I had about 1500 calories in the front and side pockets. I had a first aid kit - all custom bits with a pair of scissors and a wide variety of adhesive tape materials from bandaids to moleskin to deal with cuts, chaffing and blisters. I had a bag of bottles - a Katadyn BeFree bottle with a carbon filter, and 2 other 1 liter bottles and one 650 mil flask. All of these bottles have the Hydrapak lids that have the open/close valve function (not sure why every flask doesn’t have these now). My two bottles for the front of the vest were the UD 500 ml flasks, because nothing else really fits in the vest but those flasks (they need to be short squatty vs long skinny). I started out with the front flasks full of water and about 500 ml in the 650 ml flask with Scratch lemon-lime. For shoes I wore the Brooks Cascadia Pro (a mistake I’ll talk about later), Darn Tough socks, Path Projects liner shorts (which also failed but in an expected way but just unfortunate timing - more on that later too), a pair of Patagonia running shorts with the liner cut out (since they always wear out), a Naked running belt, a short sleeve light weight wool shirt, OR Sleeves, a Black Diamond Distance LT Hat that is light and breathable with mesh side panels and a pair of Ombraz sunglasses (which solve a problem that I never really had a problem with : sunglasses with side arms). And over that I wore a Patagonia Houdini to keep me warm, by far my most used/favorite bit of kit (as the Brits would say). My watch was a Coros Vertix and I wore the Coros heart rate monitor on my arm.

I also had 4 cameras - 2 film cameras - small compacts - Minolta TC-1 with a roll of Kodak Porta 800, and Olympus Stylus Epic with a roll of Kodak Porta 400. The always versatile Sony RX100VII - that has the best quality to size ratio of a point and shoot (f/2.8 24-200mm lens - I just wish they’d come out with the RX100VII already its been since 2019!) and finally a GoPro 13 on a flexi pod. I had both the GoPro, the RX100 and the 650 ml bottle stuffed in my running belt and one film camera was in a front pouch and one was in the vest back. (Started out at ISO 800 for lower light, and then switched to ISO 400 during the day). I kept my iPhone in my vest’s side pouch and I also had a Garmin Mini that was tracking my position and that I could send satellite text messages to my wife. And I carried with me two foldable Black Diamond Z-poles.

Lastly - I had 3600 calories in the form of gels and bars. As noted I started out with about half in vest pouches and side pockets and had the other half stuffed in a sack in the bottom of the pack. My favorites are the Salty Peanut Power Gel, then the Huma gels. I like the Quantum Energy Squares because they are quite salty/savory vs. being too sweet. I had some Trail Butters and a couple of peanut Enervit C2:1PRO bars (which I did not like) and 2 big Precision Fuel PF 90 Gels which had 360 carbs. I found the one humongous Precision gel I ate to be absolutely disgusting; they are supposed to be “palate fatigue” resistant by not having a flavor but they tasted slightly sweet and had the consistency of starchy runny glue, probably didn’t help that when I took a hit - I squeezed a HUGE mouthful in but I gagged it down. I never got to the other one as it dropped to the bottom of the list - and it was so big it triggered a bag search at airport security coming home (but not going?) and I abandoned that gel to the garbage bin. I also had 6 bags of 1.5 scoops of lemon lime Scratch Hydration mix with 120 calories) and I ended up consuming 5 of those bags. (More on water later).

Thoughts as we set out

Seth had expressed some low grade anxiety about how he’d do on the distance - since this would be the farthest he’d ever gone; though he’d been training by running consistently before work - getting in the vert - though not as many long days due to work and life commitments.

I was pushing right up to the brink of cancelling. I retired at the end of January - and with the extra time I’d been chopping some huge Doug Fir rounds, hauling them into place and I cut down a 25 foot - 6 limbed large bush/tree that was always dumping leaves everywhere and hauled the branches to the back forty (they were really heavy due to all the water the wood carried) combined with the usual scrambling over blow downs, walking on fallen logs and jumping off. In the midst of all of that I strained my right quad - just above the knee - and it felt like a dull ache deep in my quad - when going downhill or doing up hill running with explosive push off. This took a long time to rehab - like 6+ weeks of PT and biking and my last test before going to Utah was just 2 different 60 minute runs on Monday and Friday in between biking to make sure I could run pain free. Those had been a success but I was nervous about whether or not the quad would hold up. I also had an issue - that I had off and on for years - where I get a nerve pain in the metatarsal of my left foot (the foot pad just behind the toes) and it lately had been bugging me again and I had mostly (re)figured out that I was irritating it because I had a tendency to tighten my shoes and as my feet swelled it pinched my toes and caused this nerve pain. So I wasn’t coming in exactly 100% healthy. And when your coming off an injury - your hyper focused on the injury area. Always checking in with your body to see what’s happening there. Is that pain? Is that phantom pain? Is that ache indicative of a problem?

In a spirit of inspiration I’d written out this mantra and sent to Seth :

We are blessed to be here; to be present for this moment; to be allowed and to be able to traverse this beautiful canyon country. May our hearts lift us, May our feet carry us; and with eyes up may we take in the world around us as we pass through. Start easy; finish strong. Forward is a pace in this beautiful place.

Section 1 : Lees Pass to La Verkin Creek (Mile 1-7)

We got our packs together at the car at the trail head and set out around 6:30 am. The air was cold but just enough to chill your fingers. The trail starts out by dropping down from the road about 500 feet to Timber Creek, and then continues to follow this creek and descend another 500 feet, until at around 4 miles the trail bends around the large Gregory butte and starts a gradual ascent up the La Verkin Creek. Both of these trails are classic southern Utah desert canyons where water is plentiful : creek bottoms lined with grasses and full of rushes and bushes and dotted with large Cottonwoods - which were bright green in the spring and full of morning light dappling through the leaves.

At mile 7, just past the side trail to the Kolob Arch view point - the trail splits. The La Verkin creek trail continues up the creek on the left (east side) but you cross the creek and this starts the Hop Valley Trail - and here you find a consistent “good water” source : Beatty Spring.

Considerations of Water

Water in the desert is always a consideration - usually it is about whether there is water at all - and then if the water is there how to ensure that it doesn’t make you sick. Most illness from water comes from bacteria. In Zion, there is a new issue that is (impossible?) to get rid of. In many of Zion’s streams there is a bacteria that grows on the riverbeds - this is kind of like the pond scum you see growing on top of water but this is a benthic cyanobacteria. Benthic means “bottom-dwelling.” and cyano means dark blue. These bacteria don’t float but live in a “carpet” attached to the riverbed, rocks, pebbles, and submerged logs. The water flowing over them can look crystal clear. And the bacteria is sometimes green, but often velvety brown, tan, or black crusts. People often mistake this bacteria for harmless river silt or dark mineral deposits on the rocks. The issue is that its not the bacteria itself that makes you sick but that it produces a nuerotoxin (Anatoxin-a) - which is a chemical - not a living organism. They are in a family of Cyanotoxin’s similar to the venom in snakes, spiders, or scorpions. Neurotoxins, attack the nervous system by binding to your nerve receptors, but unlike good neurotransmitters they don’t “let go.” This causes permanent “on” signals to your muscles - essentially ‘hijacking’ the communication between your brain and your muscles. Resulting in tingling, numbness, muscle twitching, slurred speech, and in extreme cases, respiratory failure.

Treatment - There is none (not portable)

This neurotoxin (anatoxin-a) is not a bacteria - it is a molecule - i.e. chemical - it dissolves in water. Standard filters have 0.1-micron holes and are designed to catch bacteria (which are about 1-5 microns). The anatoxin-a is a molecule, approximately 0.0008 to 0.001 microns or a 1,000 times smaller than the pores in those filters.

The only way to remove these neurotoxins from water, is not to filter (which is mechanical) but instead you need to chemically destroy the molecule or “absorb” it. One “absorption” method is using activated carbon, which a high-grade carbon that has been “activated” by high heat, which creates millions of tiny internal pores. This gives activated carbon a massive surface area—one gram of activated carbon has the surface area of several football fields. Activated carbon works through adsorption (with a d), a process where toxin molecules chemically bind to the carbon’s surface like lint to Velcro. However, this is not an instant process. It requires significant contact time (often 10–15 minutes) for the molecules to find and latch onto a binding site. Portable water filters, which quickly force water through small carbon filters lack the necessary contact time and carbon volume to reliably remove dissolved neurotoxins like Anatoxin-a. Commercial water treatment plants let water slowly seep through large activated carbon for 10-15 minutes.

And don’t mistakenly think that boiling will help - which doesn’t “kill it” because its not alive, its not a bacteria nor does it breakdown the neurotoxin - it just concentrates it in the water - making it more dangerous. Same goes for chlorine tablets - which again kill bacteria, but aren’t strong enough to destroy cyanotoxins at standard doses. Municipal plants use much higher concentrations and longer “dwell times” to achieve this.

Don’t drink the water - there is toxin in the water

So that means you need to find reliable clean water that is not infected with cyanobacteria - you should always consult the Park’s website for current reports - (big floods and water flow can remove the cyanobacteria and result in safe water conditions).

But back to the report - Beatty Creek is across the creek on the right hand side of Hop Valley trail - seeping out of the hillside right at mile 7 and is often the last good “guaranteed” water source until Cabin Spring at mile 32.5 (25 miles away!) - there are a few other springs that may or may not be flowing depending on conditions : Wildcat Springs at mile 21, Potato Hollow at mile 28 or, Sawmill Springs (off a side trail) at mile 23.5. We drank our fill at Beatty and filled up 3 liters each. I remixed another liter with Scratch hydration mix. And then we started up the Hop Valley Trail.

Hop Valley - 7 Miles of Beauty and Soul Sucking Sand (Mile 7-14)

The Hop Valley trail starts out with a 1 mile punchy 600 ft climb to the top of the ridge at 5900 ft above sea level. Since I live at 600 feet - on any steep climb I would feel the elevation and my breathing would become more ragged. A short half-mile descent brings you to the sandy bottom of Hop Valley—so named for the hops once grown here. The valley opens into a broad, silty wash framed by the cliff walls of Burnt and Langston Mountain that tower 600–1,000 feet above you. In the spring time the shallow Hop Creek meandered through the river bottom and the trail ends up crossing it half a dozen times. The water is shallow enough and or the creek not wide enough that for the most part we were able to keep our feet dry. The shallow water reflected the morning light of the towering red cliffs among the valley’s green grassy bluffs. The sand makes for slow going - it steals your energy - you can’t push off and running seems a pointless exercise in futility so we just tromped along.

The Hop Valley trail passes in and out of private ranch land, and in one of the private areas there was a bench that said : “This Chair Was Placed In Memory Of Corey Ray Buxton. For All That Need A Rest July 31, 2012” - so we sat down and Seth gave me a Trader Joes crustless PB&J and we took a short break to get some more calories in.

Checking in on my body - at this point - my left foot was starting to have the metatarsal nerve pain - that felt like shooting pain - but it was only like a 3 out of 10. And around mile 11 I started to feel some hot spot chaffing in between my thunder thighs. At mile 14 we reached the Hop Valley trail head and a bathroom and a road! (This would be a place where you could cache water if you got someone to make the 13 mile drive up from Virgin Utah). We stopped to use the facilities - and I loosened my left shoe - which immediately provided relief from the pain. A bathroom examination revealed that my chaffing was the result of a small hole starting to wear through my Path Projects liners - which was causing a hot spot. I wear these liners every day on my runs and wear and tear is a normal part of usage. Digging into my trusty first aid kit I cut off a long piece of Medical Surgical Tape - which is thin and soft and has an adhesive backing and I applied it to the inside of my thigh; I cut an extra piece of tape in case I had to reapply (I did once more with Seth’s help) and that pretty much solved the chaffing issue. Seth applied a couple of patches to his heels and we carried on.

Wildcat Canyon to White Rim (Mile 14-22.5)

Immediately after you cross the road the trail switches to a short bit of slick rock before you cross over a broad grassy field on a 3 mile connector trail that drops you into the Wildcat Canyon trail. This was one of my favorite sections of the trail that passes through a beautiful mature ponderosa pine forest with beautiful green grass. There is a gradual climb of 600 vertical feet over 3 miles. Somewhere along the climb as we passed through the forest - my foot continued to cramp with nerve pain that shot up to 3 to 5 out of 10 and I sat down to loosen my shoe again and to tape up my thigh to stave off the chaffing. Seth had some boiled potatoes and gave me one. I told him I would start walking - as his walking pace is faster than mine and I knew he’d catch up.

The trail then starts to traverse a deep canyon above Blue Creek and just as you loop around the end of the canyon to start the climb up to the White Rim - at mile 21 there is a small stand of aspens and a gurgling spring flowing between the rocks. This is Wildcat Springs, the ranger had informed us when Seth picked up our permit that the spring was flowing. Again we drank our fill - and filled up our bottles, this time I only took two liters (1 being full of Scratch). Seth caught up to me here and again I left him to refill while I tromped up the last mile to the junction of the White Rim trail - this was our high point of elevation 7400 ft above sea level.

White Rim to The Lodge (Mile 22.5-38.5)

The first four miles is on the rim - but since your set back from the edge there are no real views, then you drop down and then up and down the high ridge - finally at 30 miles you are really on the edge of the rim and you have views off into the distant white topped canyons from whence the rim gets its name. In this section I had loosened my shoe a couple of times as the left foot nerve pain flared up. Eventually the pain spiked to 6 and then 7 and sometimes 8/9 - enough that I audibly groaned with each step. I had had tired hurt feet before - in particular on my Issy Alps 100, but this was by far the worst pain I’d ever experienced on the trail. I stopped and loosened my shoe one last time - to where it was almost entirely just open near the toes and sinched at the top of the tongue. Eventually the pain subsided to just a 2 or 3 out of 10 but with a couple of steep climbs and the pain - I was reduced to a repetitive focused state of step, poles, step poles, ragged breath : over and over again. Nothing else mattered; no other thought was present but just that and I’d often find myself just staring at the ground as I marched along. Every once and a while I would break out of my zone and remember my mantra and lift my eyes and take in the scenery around me. The top of the rim was a mix of large ponderosa pines and open plains of manzanita bushes that were dotted by dark spires of pines shattered and burnt by lightning. At times I felt flushed and like if I laid down that I’d break out into a fever and sweats, but I did not stop and carried on.

Eventually as we got towards the end of the rim and neared Cabin Spring (mile 32.7) my feet (both) were now just hurting from the long day - like a 2 or 3 out of 10. Even if we’d had gas in the tank to try to run at all along the rim, the wheels had fallen off. We took the wrong side trail for a short bit and I noticed the spring was down the hill and we had to back track and we were just so tired and ready to be done. Seth was not in the mood and did not want to stop. I was out of water but he had a liter left and gave me half.

It was now 7:30 pm. When we started 13 hours before at 6:30 am - we optimistically thought that we’d be done for sure by 6:30 pm, but the sand, the fatigue and the stops at springs and sitting down for some food had put us on the edge of the White Rim well past 6:30. We had only 5 miles to go - which dropped 2400 vertical feet. Under “normal” conditions - with gravity on your side - IF you have legs left you can descend at a faster pace. However by this time Seth’s legs were not up descending, each step down he’d groan as his knees gave out with each step. We made the dramatic descent down the White Rim and then traversed toward Scouts Lookout - just below Angels Landing. The sun went down at 8:18 and by 9pm the darkness over took us.

Of course we had left our headlamps at the car. So we pulled out our phones and turned on their flashlights. As we passed Scouts Lookout the trail turned to pavement - and the very steep Walters Wiggles zig zags straight down the cliff face with 21 switch backs that loose 260 feet in .2 miles for a 25% grade. This grade really caused Seth to groan with each step. He later said he wanted to stop and yell swear words at the trail. As we descended I had developed a raspy cough, likely from all the ragged breathing. We were worn down to the bone. When you are pushed to what you think is the limit and yet the only option is forward. There is no quitting, no DNF because there is nowhere to go; no one to pick you up, no helicopter to lift you out, no zip line to to take you down because your in the middle of the wilderness. The only way out is through. And when literally every foot step hurts and in the darkness the trail goes on and on - you just have to keep putting one foot in front of another and just keep going until you reach the end.

We descended down the trail cut out of the cliff - and the nearly full moon lit up the valley below. We finally reached the river - and then finally we reached the bridge - that crossed the Virgin river and trail was done. We were on the road and yet we still had ¾ of a mile to get to the Lodge. We arrived finally at 10:15 pm. The restaurant had closed at 9 pm. But it didn’t really matter because a well executed ultra is not only a feat of ultra endurance but also ultra eating. When was the last time you ate for 15 hours straight? By the end I had sores on my tongue from all the sugar I’d consumed over the course of 3700 calories. I wasn’t really hungry. At the front desk attendant got our bags and checked us in - I went to the drinking fountain and sucked down a lot of water. We trundled off to our room and I sat down in the shower and washed off the sweat and the grime. I got under the covers that had been jack hammered under the mattress and after tearing at them to give my tired feet some room I fell asleep hard and didn’t wake up once to go pee - a sure sign that I was still dehydrated.

Lessons Learned & The Aftermath

First of all - thankfully - my quad didn’t bother me at all - so that injury ended up being “done” - though since returning home on the 4th day I went on a short recovery run and my quad had twinges again - so who knows. But on the traverse - the quad was fine.

Secondly - I was happy how food and water went.

Thirdly - we did run the flats and downs the first 7 miles - (until we hit sand) and I think we would have been better off if we had just started out with a brisk walk and walked the entire time - not pounded the legs with even easy running.

Fourth : I think wearing the Brooks Cascadia Pro was a mistake - they have a carbon fiber plate in them - which is designed when you run to spring back and give you an energy return - but when you walk 90% of the time - I think that plate exacerbated my already irritated metatarsal and contributed to the nerve pain. On that recovery run on day 4 - about 40 minutes into the run my left foot started hurting again - so it’s still irritated. Need to let it rehab and figure out how to strengthen and get it back to healthy.

Lastly : despite the pain and the fatigue - the trip was really great. The weather was perfect, the scenery was beautiful and company was grand. And 3 days after when Seth and I were talking we both said we’d like to give it another go. ;)

Strava route : https://www.strava.com/activities/18301428737

Photos on Flickr : https://flic.kr/s/aHBqjCSNPw

I have consistently run 30-40 miles a week for the past 5+ years, and been staying active for more than 22 years. This is more about time on my feet (10-12 hours) going at my moderate old man pace with vertical gain thrown in. (Not to brag but also to brag 2025 was 1,794.8 miles and 439,574 vertical feet). This was in part how I survived Amazon - it was my therapy, my stress relief and my way of not just keeping me in shape but giving me something outside of work to focus on. And I’ve been quite disciplined - always run in the morning before work. Get up and go. Have a 7 am meeting and 90 minute run on the schedule? Well just have to get up at 4:30 am, and start running by 5 am in order to shower and dial in for the 7 am meeting from home.

I didn’t expect anything to change when I retired - and it didn’t I stayed in my routine and followed my weekly plan. Though I did shift from getting up at 4:30 am to getting up at like 7 am ;) But I did have more time on my hands. I have a bunch of projects (like organizing 100s of thousands of digital files and rebuilding my website with Claude Code). And that meant that I could also do some “chores” around the house.

Two of those activities were splitting big Doug Fir rounds for firewood. This involved moving heavy rounds in place, splitting them with a maul and a sledge hammer and then splitting them into smaller pieces with an axe and then stacking. Another was me deciding to cut down a Japanese Spindletree that grew next to our house. The Spindletree is really a bush and it had grown 20+ feet tall and it was continually dropping leaves and I found it annoying. So I cut it down.

See it has like 6 trunks? And that wood was super dense, wet and full of water. I cut it down mostly with my electric chainsaw (which is not really suited for this task) and then I cut off the branches and carried very heavy logs (sometimes with Miles help) down to the back forty for them to decompose.

And then on one run I also jumped off of a downed tree on Old Griz and hit the ground.

Somewhere in between all of that I strained my right quad. When I ran uphill or downhill I could feel a pain deep in my right quad about 4 inches above my knee. And then as part of the healing / injury process my knee swolled up with extra fluid and I couldn’t bend my knee. And that has been about a little over a month now - and things are very slowly slowly getting better. My bike ride in the glorious 70 degree sunshine yesterday felt great and today’s run felt pretty ok.

So I am sure that eventually things will get back to normal; but a good lesson that having more time on your hands and doing more might not always be the best thing. ;) And talking to Koop - he said this is not the first time he has had this happen. So be on the lookout how you change up your training and activities.

You’ve all heard the maxim : “The best camera is the one you have with you.” - but I’ve always liked to reply to people who ask me - “Do you always carry a camera?” : “Its hard to take a photo if you don’t have a camera with you.” - and part of that is my belief that despite the camera that most people have with them : phone cameras - there are still yet better cameras that you can carry with you to make better photographs. Cameras that have more resolution (yes it does matter, and no your phone camera does NOT take 48megapixel photos) and that have options with lenses.

Like most things in life a system is a set of compromises and trade offs, even the basics of a camera itself are a trade off between ISO, Shutter Speed and Apeture. With digital cameras there are trade offs in size of the sensor, size of the body, mirror or mirrorless, fixed lens or removable lenses and other factors that impact frames per second you can shoot, like the camera’s internal memory buffer and the write speed of the SD cards you use. But…..

Sometimes in this digital age of immediacy it is nice to slow things down and using the analog antecedents : film and film cameras. They force you to slow down and teach you things about the various balancing trade offs in photography - when you only have 24 or 36 photos you can take - it makes you stop and consider things more. Joe Grant, has been one of my inspirations about shooting film. He has written about this in 2018 on iRunFar.

I needed to learn how to look, to observe, to pay attention to a scene, emphasizing the importance of composition. … I needed to practice… [to] … spend time learning the craft, underlining the value of the process.

I also been inspired by Seth always carrying a film camera on our adventures. I once did the Foss Lakes Traverse with just a film camera as my main camera. And have an idea banging around in my head to shoot a trail race entirely on film. I’ve been buying some (more) old film cameras from eBay and when I mentioned my film race idea to Ryan Thrower, he said he plans to shoot a roll a day on Mammoth this year and he asked for some advice on what film camera to use.

So…. here is my take and advice.

First of all decide between fixed lens point and shoot or standard film camera with removable lenses. Assuming you don’t want to really get into the range finder era, and stick with standard 35MM - I think the lightest and best option is the Olympus OM-2 - which has an aperture priority mode - and you could put on a 50mm f/1.8 or a 35mm f/2.8 - but then you have to focus and your not really getting something that is pocketable.

Here is an Olympus OM-2 with a 50M f/1.8. They run around $200-300 on eBay.

There are other “famous” brands in this same category - Nikon and Canon and Pentax and Minolta but they all are larger.

Then you have a myriad of true point and shoots. I think this what you want because this is truly “pocketable” - its a fixed lens with auto focus. Here are 3 that I own.

Pentax PC35 - 35mm f/2.8 - they run about $120-150 on eBay.

This is a Minolta TC-1 28mm f/3.5 - they run about $900-1100 on eBay

And finally my personal favorite the Olympus StylusEpic DLX 35mm f/2.8 which runs about $250-350 on eBay.

It’s teeny bit longer that the Minolta - but is thinner (and less expensive) and due to its shape seems to slide in and out of my pocket better, and the manual slide open feels a bit more robust that the beautiful magical electronic shutter on the Minolta.

The other two things to consider are, you really want a “slide open shutter” - nothing with a lens cap - as that is just gonna get lost and risk getting your lens dirty. And finally remember there is a lot of moisture when running - water from the elements, sweat from exertion - so you want to have a way to keep the camera dry. I have the original soft case for the Olympus and then keep it in ziploc.

Here are a few more pics from that day in the Alpine Lakes :

First - to clarify a few things. AI is a very broad term that is not often used clearly. We have been doing machine-learning - big-data systems for 15+ years and employing those in tech to solve problems at scale that were not possible before ML. A lot of people call this ML tech AI. But really today AI is being applied to Large Language Models- that have been employed most popularly in Chat Bots like ChatGPT, Gemini and Claude. The early versions of those Chat Bots were impressive because you could use natural language to ask questions and they had ingested huge amounts (all?) of data and had the computer/synthetic neural network ability to produce output that was “natural human talking.” - And they were often tuned to be “nice” and sycophantic and complimentary - especially ChatGPT because the goal was interaction/engagement and usage. But they were (and still are) often wrong about what they produce - because they are based on statistical probabilistic reasoning instead of based on a world model that is anchored in “reality”/physics. They weren’t good at understanding time and were not good at doing math for example. And most of all - there were no sentient, often referred to as Artificial General Intelligence (AGI). Zealots have acclaimed that the Chat Bots were alive - and/or that it was coming soon, always in the next 6 months or the next version. The doomsayers predict if/when AI does attain AGI it will destroy humanity like Skynet in Terminator or “The Machines” in the Matrix. I don’t believe that AI is anywhere close to sentient - though I do think we need protective regulations for how AI is used in society.

Second - in the last 6 months as Chat Bots have evolved - with different companies building them - Google → Gemini, OpenAI → ChatGPT and Anthropic → Claude - they have gotten more useful. Some of this is a a bit of a self fulfilling prophecy for me - because I didn’t use them a lot at Amazon - first because of the restrictions on their usage - not disseminating confidential Amazon information outside of the corporate network. And secondly because most of what I did at Amazon relied on my knowledge built up over 21 years, my judgement and my ability to write in as an Amazonian - and no external AI system was going to be able to help there. AND in 2025 - frankly the internal approved Amazon chat bots and systems were not really that good (at least in my usage - both in terms of accuracy and quality of output). But since I’ve been retired - I have more time on my hands - and I have some needs that had me trying out AI.

I have been using Gemini generally for all kinds of technical troubleshooting and investigation. For example - I replaced my Mac Mini with a Mac Studio - and I have a Synology DS918 - which is a Network Attached Storage (NAS) with Redundant Array of Independent Disks setup (RAID) that increases reliability because data is written to multiple drives so that if any one fails you don’t lose anything. Anyway I had my Synology hard wired with an Ethernet cable to my Mac Studio - but it couldn’t be seen. And Gemini walked me through an entire series of troublehshooting steps - changing the subnet / mask to isolate that Ethernet was working - the cable was plugged in but no data was being transferred. We reset the Synology, un-plugged and plugged back in the cable. And it “figured out” that the likely result was the fact that Mac Studio has a 10Gb Ethernet controller and the Synology has a 1GB controller, and I was using a Cat5e cable, and that can’t reliably handle those speeds and because Cat5e cables has less internal twisting and shielding and the 10Gb controllers are much more sensitive to crosstalk and electromagnetic interference such that the Mac Studio often detects “noise” on the line and simply refuses to communicate to protect against data corruption causing the connection to constantly drop. So Gemini suggested to go buy a Cat6a cable which has metal shielding to prevent interference and we reset the network to static settings and voila I could connect to the Synology again.

I could have figured this out by the usual method - lots of googling, lots of reading forums, maybe I’d find a post by a random tech dude who documented what they’d done to fix a similar problem - or I’d have to write Synology support and wait. But Gemini had slurped all that data and synthesized an understanding of how it all fit together and provided the solution, after going through the appropriate troubleshooting steps. Along the way I could upload pictures of the back of my Synology and screen captures of my Mac Studio configuration and it could understand them all and recommend next steps.

And one of the major retirement projects that I had on my plate was to clean up my digital files. I have been shooting digital photos since 1999 when Sofi was born, (I bought a Sony Macvia camera that wrote pictures to a 3.5 inch floppy disc) and I bought a DSLR in 2005 when I started at Amazon (Nikon D70). In 2023 I kept 19,377 photos, and likely took double or triple that amount, In 2025 I kept 17,570 photos. I have a LOT of digital files - all the original “negatives” as well as all the exported files - not to mention the gigabytes and gigabytes of video footage that I have shot over the years. And in the last 5 years my digital file management has been more emergency based than strategically and rigidly disciplined. Meaning that whenever my MacBook ran out of hard disk space - I would move those files off of my laptop. And I would always copy/backup at the end of year - when I’d start a new annual Lightroom Catalog and export last years catalog. I’d move them on to an external SSD, sometimes I’d copy those to a series of G-RAID drives, sometimes to Synology but not always. And since often it had been a year since I’d last really spent time on this - I wasn’t always consistent and/or couldn’t remember from back up episode to episode where exactly I had stored files. This left me with my digital files in a mess, not all in a consistent location, with multiple copies in multiple places across many different drives. I had dreamed of hiring some bright high school student with technical aptitude, digital file knowledge and attention to detail as a digital clean up assistant to help me.

I had another sad event happen - where I had a 4TB SSD that I had left in exFAT format that had “gone bad” - (exFAT is susceptible data loss if the disc is unplugged while recording the map of file locations) and I installed Claude Code to help me attempt to recover files. Wow - I was blown away by Claude Code’s capabilities. Claude Code is a new coding specific AI model that uses the Claude Chat Bot to let you chat with in natural language, describe your tehcnical problems and goals and then it will go off and write scripts or programs to accomplish your goals. This is a technical tool for developers or technical folks - it uses the terminal for input and output. Claude Code walked through a long series of troubleshooting steps to try to recover my files - sadly to no avail. But then I also started using Claude Code to help me analyze all of my hard drives for the file contents and to help organize my files. Claude Code is my digital assistant - it has context and remembers things from day to day and copy, verifies, cleans up and advises. For example :

and this

I also have another project in Claude Code where I am having it help me rebuild my website using the FlickrAPI - still a work in progress. And I am also having Gemini teach me how to do animation in Procreate Dreams.

So I am over my skepticism on AI - it is useful. It is having an impact on me and my productivity, but it is not replacing me - it is helping me - get stuff done and be much much more efficient.

I’m now 3 weeks out from retirement. My last post was about the immediate silence in communication that happened for me overnight. This is also about silence and cessation but in a different way.

At Amazon there was in my role an unrelenting incessant sense of something always being the next priority. Something that always next needing to be done. And almost always more than one thing at a time. The next QBR, or a 3 year plan or OP1 or Q4 prep or a deep dive into why CountryX isn’t growing in segmentZ or reviewing the write up for a Jassy escalation or figuring out what to cut from the roadmap due to attrition. On and on and on, a continuously swirling set of #1 priorities. And in my role I had to keep track of and make sure to push forward these priorities with the team. This sense of “most important” occupies a huge area of the front part of your brain. I would consistently check in and review what’s important and think about it and mentally noodle and attempt to solve or gestate ideas about these things.

My morning runs while working were a huge part of me giving my body and my headspace a break from this incessant buzz. Sometimes though even these work priorities would worm their way into my thoughts while I ran.

But today in my morning snowy run in Logan - I was able to mentally step back and notice how nice it’s been to not have that incessant priority any more Amazon in my brain. The mental clarity and freedom has been a breath of fresh air.

The quiet of that work prioritization is very nice and peaceful and I enjoy that my time and my mental energy is my own.

When the volume is always turned up and you’re part of a cog machine that is moving fast, you don’t know what it’s gonna be like, stepping away from that until your gone.

Three days into retirement - this was part of my initial reaction. Different from vacation - because even on vacation you knew the machinery still existed. And even if you tried not to look at email or work chat - you’d steal a view every now and again just to check up on things. And often in emergent situations - people would still reach out.

But this feels different, even after just three days removed, the silence, and the lack of communication was pretty deafening. It made me wonder whether all communication is transactional, and whether the reason people communicate with each other or communicated with me, was because they had to, because they were part of the machine that I was part of; because I was the boss. It’s a shock to the system to go from moving that fast and having that many threads of communication and always being asked something : to then nothing.

I am sure things will become a new normal over time - but that’s the feeling right now.

As I get ready to retire from Amazon - I wanted to take the opportunity to look back on my time here (20 years, 11 months) as well as my career generally, call out some key career transitions and offer a few thoughts about what Amazon has meant to me, what I have learned, how and why I’ve been here as long as I have and my advice for those that continue.

And if you think this is too long - just use AI to summarize it for you. :)

As I’ve contemplated “my when to leave” - I’ve had this letter open in a file for a while and I’ve come back to it time and again to write this out. I’ve wanted to make sure that I tried to capture my thoughts accurately from all the angles.

First off - Amazon is stage 3 of my tech career that began during graduate school. I never meant to get into tech - I was originally heavily engaged in trying to figure out what to do with my under graduate degree in history. I often felt lost early in my mid 20s when I was struggling to figure out what I’d be when I grew up. Originally I had an idea to be a professor - to get a PHD in History and to teach…. But after returning from Taiwan, where I’d studied Chinese and finally got my BA in History via independent study in 1993 - and after deciding not to join the military or take the manager job at Sbarro’s in the Mall - I went back to graduate school. I loved my graduate program but I was recognizing (albeit slowly) that a PhD was going to be another 6 years and that professorship might be tough to get a job that would be able to support my family.

And then I got completely lucky and fell into technology. While in the graduate program in Asian Studies I was an assistant to Dr. William J. Hamblin, and then the World Wide Web launched (1995) and Hamblin asked me to build a website with clickable image maps showing the growth of the Assyrian Empire. That started me down the path of figuring out how to build websites, which led me to my first “tech job”at Nu Skin in Provo Utah and that continued a pattern that has been consistent my entire life : curiosity, learn something new, do something new and keep going. I have always been driven by a creative desire to understand the world around me, a desire to learn - from reading the Encyclopedia as a child, or browsing many thin yellow volumes of the National Geographic : this curiosity persisted throughout my work. Maybe my attitude, my effort and my energy helped but I didn’t plan to end up here - though I’m grateful for everything I’ve had the opportunity to do. From Nu Skin I got a job at the startup WebLogic in San Francisco (a separate interesting journey working with some of the brightest minds in Java and distributed computing). I continued with BEA Systems (that acquired WebLogic) for 6 more years and we eventually moved with them up to Seattle.

We moved to Issaquah outside Seattle in April of 2004; when I returned from my first backpacking trip in August I got a voice mail telling me that BEA was closing the Seattle office and I started interviewing. Brent Curtis got me an introduction at Amazon - I was late for my coffee chat with Charlie Bell at Pac Med - I must have passed because I met my hiring manager Andrew Certain at Starbucks in Factoria - where he handed me something called a PRFAQ - a doc that described a new product offering at Amazon that he owned called Self Service Ordering and Fulfillment (SSOF). Andrew arranged an interview - Paul Kotas was my bar raiser. I was offered two roles - both Technical Program Manager (TPM) positions. One was working for Andrew on SSOF and the other was working for Tim Craycroft who ran the Program Office - where I’d be working on a large cross functional program to deprecate the Single Detail Page merchants program that sold BMVD and merge it with the more “modern” M@ Merchants system. I was leaning towards the latter as it sounded challenging but Charlie Bell suggested that I take the role working for Andrew on SSOF.

I started at Amazon in Feb of 2005. I was hire # 3 on the Self Service Ordering and Fulfillment (SSOF) team - my 2nd day at Amazon I was in a Jeff Bezos meeting where we reviewed the state of SSOF. Jeff wanted us to go faster. They still hadn’t started building anything - I was supposed to help with that. The PM (Josh Sandbulte) quit Amazon a few weeks after I started, and I recall him stating he was glad that he’d never have to meet with Jeff again. (Josh seemed like he was stuck in analysis paralysis - something Jeff did not tolerate). I was initially assigned to help launch a new Merchant : Bebe on Amazon’s dot.com systems. Amazon would run the website and would also handle the fulfillment of orders. No one had done a fulfillment program since Target launched as a website on Amazon systems with fulfillment in 2002. My job was to figure out how to make fulfillment work for Bebe and then take that learning and apply it to SSOF. In July, 5 months after I started my hiring manager (Andrew Certain) took a leave of absence. That left me as the sole team member of SSOF, and then Tom Taylor was assigned to own the SSOF program. He brought in a Principal TPM - Chad Goelzer to run fulfillment Ops, Principal Engineer John Chenault to figure out systems. I was to leverage my Bebe learnings and go figure out how to make ordering work. Tom quickly hired a 3rd TPM - Dipika Kapadia to work on Seller experience and Joe Walowski as our first PM. Tom prioritized, he simplified, he cut features and focused us in a singular direction : to launch by 12/31/2005. We had 5 sellers to start. I drove out to Issaquah to label and load books of one of those sellers (Grassroots Books) to ship to RNO1. We launched in April of 2006 - and we quickly renamed the program to Fulfillment by Amazon.

FBA started off super small. We quickly signed up another 15 sellers but there were a lot of missing features (no inventory view, no view of orders, no returns, no fees etc etc). So we paused while we built some key features. At the end of 2006, Bezos had us set a goal for “# of merchants we would launch on FBA in 2007.” I think we came up with 1500. Jeff came back that we should do 25,000! That was a lesson to focus on self service. Jeff emphasized that a seller should be able to come to the Seller website, sign up, read the docs, and use the seller portal to create an inbound shipment and send in inventory to Amazon’s warehouse all without ever talking to a person. We missed the 25K, but we hit our stride on self service and as Prime launched and FBA inventory qualified to be branded as Prime, the program really took off.

I was promoted to Principal TPM in Oct of 2006 - I remember being shocked because Tom just called me into my office and informed me. We’d never discussed career development - I guess he thought I was quickly working at that level after joining and helping launch FBA. In some ways this messed up my mental model for career progression - as I expected my next promotion would take place the same way : it would just happen. I was mistaken - more on that later.

For the next 3-4 years I kept on taking on large complicated projects within the FBA space. Launching FBA in UK and Japan. Launching Multi Channel Fulfillment (MCF : ability to tell Amazon to ship an item of your inventory in our warehouse to any address - typically because it sold on another website), Launching Joyo.com in China integrated with Amazon FC systems via MCF. I resisted being a manager - even though I recognized later that they tried. They had me hire a writer early on to work on FBA Seller help docs - he reported to me. But he was a bit odd - I guess he had back problems - so he brought in a camping sleeping pad, and screwed his monitor to the bottom of a door desk and would lay underneath on the pad flat on his back while staring up at the screen and typing on keyboard. Not that someone couldn’t produce great results with that setup - but he did not produce great docs in copious amounts and was caught sleeping a couple of times. And I couldn’t get up the courage to give him feedback on his performance. Someone else had to take over helping him find employment elsewhere.

Eventually in 2010 - we got a new tech Director : Adrian Inglis and I moved to report to him. I had “hired” internally another L7 Principal TPM (the fabulous Lorraine Nicholson), and I was ostensibly in name a Manager of one person who didn’t require any management. Adrian said to me - “Next year we are going to make you into a real manager, none of this fake manager stuff.” - Adrian had a way of being very direct, very blunt and pushing me to grow. He probably had more to do with my career growth at Amazon than anyone else. I had another SDE Nancy Walsh transition over to being a TPM reporting to me. Then I finally convinced Eric Gregory to come work for me - we had worked together on the Bebe launch. And I convinced Ben Lee, who I had met during the Joyo launch to also come work for me on FBA. We had a small and mighty team of 4 and we got stuff done!

By this time I was committed to management - Adrian gave me a ton of feedback and I was asking him about my next transition - to L7 Software Development Manager. Adrian kept acknowledging his support but also emphasized that this was a move that required timing and planning. While the overall number of direct reports would be similar in size - the scope of the role - a manager of managers and the total number of the overall team would dramatically increase. Finally I hit upon the idea of : “Why not send Mark to Beijing to manage the software engineers there?” We had hired a bunch of engineers to work on FBA, while being situated in Beijing. After discussion in talent review and being approved by Tom and finance - me and my wife Stacey and our 2 kids (Sofi age 10 and Miles age 7) were set to move to Beijing in Feb of 2011.

What a grand adventure, both professionally and personally. 3 weeks after I arrived in Beijing, the one software manager quit and I was left with a team of about 13 engineers all reporting to me. I quickly had to hire a manager, and upgrade some of the team. When we left 18 months later, I had hired 3 managers, Wang Lei as my replacement and had built the team to around 22 engineers. I came back to Seattle to a much larger role - and under Adrian’s tutelage began to understand that moving up to Tech Director was a step change in how I operated and managed.

At Amazon we often think of levels as +1 increments, going from 5 to 6 or 7 to 8. In reality these moves are exponential jumps in responsibilities, expectations and execution. Key things that I learned in getting promoted to Director in April of 2014. First people are your product. You still deliver stuff, but more through others. You must always be hiring and developing - your ability to scale as a leader is only as good as the leaders you have on your team and them in turn. Second, managing people is messy and that means you have to roll up your sleeves, solve human interaction issues and to clean up messes. That means being willing and able to give feedback, good and constructive and deal with performance issues and the HR process of improvement or separation. Third - you have to partner and be willing to say yes, lean in and figure out how to get stuff done. Coming from a TPM / Engineering background - I was reluctant to commit before measuring twice : “Not sure if we can hit that goal, I need to get final requirements, estimate scoping and then look at our budget and let you know.” - When really what your business partners are looking for is “Yes, we will figure it out.” - or good very detailed reason to push back based on principle, data, customer feedback etc. Lastly - I learned that you had to, and were expected to toughen up, to push others to higher standards and to do more; to invent and simplify, to be creative in approaches to problems. I never had a problem pushing myself to do this - but it was a new thing to be comfortable to engage others under your direct leadership to do this. And to build mechanisms to do it at scale.

Another couple of key lessons in career that came out of FBA - first the importance of staying in a place where you can build career equity. This of course can be highly dependent on an organization that is growing with good managers who support your growth. And FBA grew like crazy. I stayed with FBA for 9 years because I had opportunities to grow as the program grew : I went from an IC to a manager, from line manager to manager of managers to Director of Software Engineering. The other key lesson, and a guiding principle of how I work (and this was re-enforced by Tom Taylor) - is that you can work hard, make history and have a lot of fun. We had a great team culture in FBA : with beer Fridays, contests (donut eating, pushups), nerf battles, lunch runs, polar plunges etc etc. What a lucky break it was to be able to join FBA - an opportunity to build a startup inside of Amazon that has grown to have such a meaningful impact on the store, attracted millions of sellers, delivered value to 100s of millions of customers; a program that has shipped 80 billion units via FBA.

6 months after my promotion to Director, I went to work with Amazon Pay (a simple way to think of Amazon Pay is Amazon’s version of Paypal) and I ran the tech teams that built systems that helped enable 3rd party payments on Sellers websites. Like FBA, I had the opportunity to learn and grow. I loved learning about the entire payments infrastructure, the “value chain” of the payments execution and the way the ecosystem worked : issuing banks, card brands, disputes and transactions. Ever wonder how “Tap and Pay” works underneath the covers? Eventually I was given the opportunity to manage new roles and learn new skills - in addition to engineering I also led data, design and product management. Two key lessons I learned at Pay - first organizational structure and composition creates unique dynamics and skews focus and priorities. In Pay, my peers were 3 sales directors and a director of marketing - in my opinion this tended to skew priorities toward closing deals or creating opportunities in market expansion through partnership or integration which overshadowed the Amazon principle of ‘working backwards’ from the customer. Secondly - I learned how challenging it is to sell a product to customers (sellers) - where your solution is not solving 100% of their needs. With Amazon Pay - we were generally an add on to Sellers websites and their existing primary payments solutions; even if we became a 100% of their payments solution we were a bolt on to their existing check out flow. Both of these created unique challenges with adoption and scaling. Finally I learned that your impact is highest when you are fully aligned with your leadership and thus I sought a new different environment and challenge. I think this is one of the best things about Amazon : the company has such a broad set of initiatives you can find new challenges and opportunities to grow and learn and Amazon encourages and welcomes people trying new things.

In looking for my next role - I wanted to get back to something where physics and moving actual stuff was involved. I naturally gravitated back towards fulfillment. I met Steve Armato at a bar after work and we chatted - he suggested I talk to Peter Larsen - who proposed a role leading MFN Fulfillment (where sellers operate the own warehouses and ship item with 3rd party carriers like UPS and FedEx) - though he said it did have some issues that would require care and special attention and focus. I recall calling Eric Broussard on a Saturday while trail running to talk to him about the opportunity. There is a circular twist of ironic fate - when I joined Amazon - my hiring manager owned Merchant Fulfillment software systems : MFN and was tasked with expanding those systems to enable Fulfillment By Amazon. I was a TPM assigned with figuring out FBA - and that program grew to transition a large part of the merchant shipments from MFN to FBA. In my opinion, this resulted in a bit of languished attention on MFN - and then 15 years later - I was to become the owner of MFN fulfillment.

I moved over in February of 2020. 10 days later Covid hit the US - we went home to work for 21 days and stayed there for roughly 3 years. MFN, which was growing in the US in low single digits annually, suddenly was growing at high double digit rates. In April 2020 MFN shipped more units than during Peak 2020. There was immense pressure to improve the MFN experience. We worked a LOT of hours. I had the “opportunity” for weekly meetings with 2 SVPs, and bi-weekly meetings with then CEO Dave Clark. We were directed by Clark to go look at an acquisition - to help sellers with fulfillment on all their sales channels - not just Amazon. In Jan of 2021 - we had our first discovery call with Matt Warren of Veeqo and in Nov - after 9 very intense long months we closed the Veeqo acquisition. Then we had the hard work of integrating Veeqo into Amazon, and adapting our fulfillment to handle off-Amazon shipments and wiring up Veeqo to use our Buy Shipping program. I learned so much throughout the process and it was a lot of fun helping a great company on-board to Amazon.

In 2025, in conjunction with a re-org - where MFN moved out of Ops transportation and integrated with External Fulfillment - I took a smaller role running just product - so that I could have more bandwidth and space to help out my wife and her health. And now the time has come to step away entirely from Amazon and start a new chapter outside of work.

Looking back I want to highlight some of key improvements the team has made in MFN fulfillment. I’ll use the US as my primary example, since I have the most familiarity, though these changes were also reflected world wide. I will try to use 2019 as a comparison to 2025. During Covid (2020) things got very skewed by the way everything shifted to ecommerce, so 2020 is not a good comp. The MFN fulfillment experience that Sellers provide for customers has a bunch of different components - how fast is the promise of delivery, what are the tools used to generate the delivery promise and what % of the promises are “fast”. Some of the key areas that had big improvements :

Handling Time - the length of time a seller takes to pick and pack the order and hand off to carrier.

In 2019 9.4% of sellers had 1D HT and in mid year 2025 1D HT configuration was 71.4%

Promise Automation - Some background : In 2019 0% of the transit times in MFN promises were automated. We used very coarse grained ranges (5-7 days) that were not precise. In 2020 we launched transit time automation, which calculated and optimized the carrier transit times for a zip5 origin to zip5 destination - for precise and accurate promises. These were much more accurate than manual transit times and were a better (faster) customer shopping experience - which produced better shopping conversion for MFN Sellers. In 2019 0% of the transit times were automated - in 2025 45% of promises were automated.

Buy Shipping - a product that lets sellers purchase shipping labels where Amazon provides negotiated rates, and predicts which shipping services will meet customer promise. In 2019 Buy Shipping US Adoption was 26.4% and by the end of 2025 that had grown to 38%.

We measure the speed or the length of the promise in days - which we internally call Click to Promise or C2P. In 2019 the C2P was 12.4 days, that is the average MFN promise was 12.4 days long and in 2025 the C2P was 8.6 days, a 30% improvement.

We also shifted more of the promises to be faster : in 2019 ~15% of the promises were between 6-8 days and in 2025 that number had increased to 45.4%

I also owned improving the Prime experience where sellers shipped Prime orders via MFN : (Seller Fulfilled Prime - SFP). One of Prime’s main benefits is FREE Two-Day Delivery, with Prime shipping speeds increasingly shipping in one-day. In 2019 the % of SFP promises at 1Day speed was 1% and 2Day speed was 14%. In 2025, 35% of SFP promises were 1D and 79.8% were 2Days. A dramatic improvement in the Prime customer shopping experience with SFP.

Taking an even wider view across everything I’ve had the opportunity to do while at Amazon : What am I most proud of? Of course I am proud of what I was able to work on and what I have learned and what “I” singular-achieved -which in the end is not really THAT much : everything at Amazon is a stacked set of dependencies and interconnecting services and individuals working as teams to produce something amazing. What makes me proud is the people I’ve had the chance to work with, the teams I’ve had the opportunity to lead and a hope that along the way I’ve made things a bit better in my fellow employees lives and through Amazon the world in general. Two projects that stand out above all the rest : 1) being involved in the launch of Fulfilment By Amazon and 2) being able to work on the acquisition of Veeqo.

How have I stayed so long? How have I “survived”? Amazon is a high demand company, not known as an easy place to work and not a place to coast. Amazon has a high hiring bar, and high expectations for how someone conducts themselves and high expectations on what people deliver. Here is my distilled advice and wisdom :

Amazon has a sense of urgency to it that can be all encompassing; As Jeff said “Speed in Business Matters” - and that has this way of creating the expectation that everything you’re working on is the most important thing and that it must be done with speed. Answer that email that comes in fwd from Jassy (rightfully) questioning an anecdote from a Seller. Figure out how to achieve the S-Team goal that you signed up for because of the importance of the program you’re working on. How can you go faster? (An oft repeated question). That is part of your job; part of why you were hired and that is the part of the force multiplier in leadership positions : The CEOs, The Senior Vice Presidents, the Vice Presidents and Directors are there to empower, focus and enable their teams to deliver.

But - if you are not careful - you can get ground down, or you can get overwhelmed with anxiety or burn out. You must be ruthless at prioritization, you must focus on what really matters and where you can make a difference. You can spend all day writing and replying to emails but not be doing the most important work. Protect your time at work; But also protect the time that is yours outside of work and the time to be with your families or significant others. Learn how to and when to let go of work; there is ALWAYS more work to be done but you should not always have to be working. If you do - you will burn out. I have been there and I have seen many people in that state. You must be careful about how much of your life you give to work and when you do so be conscious of that decision. Some people’s hobby is work; some people are defined by work - those are personal choices that are not necessarily wrong - but you should be aware of the choices you are making and why.

Build time to think. Build time to recharge. Build time to de-stress. For me that has always been the outdoors - hiking or running before work. In the last 10 years, I always run to start my day before work; no matter what time pressure I had on my morning - I’d just get up earlier to get my run in. 7 am meeting? 90 minute run on the schedule? Set the alarm for 4:30 and get out the door by 5 am and be ready to roll for that 7 am meeting. And after you have run a 100 miles in one go - you can push through a lot of tough things at work.

If you are a manager - protect your team. Care for them and their interests; don’t coddle them or let them deliver to low standards. Be genuine. Be Straightforward. Trust your instincts; dive in when necessary - the details always matter and they illuminate the way that gets lost in the aggregate or the average. My organization philosophy has always been - hire great people based on Amazon’s high standards of thought, execution and leadership skills. Give them ownership and responsibility, empower them with lofty goals and set them up with great leadership who are empowered to do the same. Audit deeply and regularly to check in on progress at all levels. CARE deeply about your employees development, success and give them feedback and support to grow and adapt and achieve. Motivate through self development, customer obsession and achievements - not through fear or by manipulation. Be fair; be equitable.

I believe that the highest ideal we can achieve is to be an inspiration for someone to live a better life and inspire them to inspire others. I have perhaps enjoyed the most developing others; helping them grow, watching them learn, achieve and excel. To develop their own skills as individuals or as managers. To see them gain experience; to give them feedback, to give them suggestions and direction; to coach them and to help them achieve success and become better employees.

Lastly, as I have mentioned before, have fun. Curiosity about what you’re doing makes everything more interesting. Learning is fun and almost everything is interesting if you go deep enough. I enjoy solving hard problems; I enjoy figuring out a way through; those that enjoy the process are more likely to succeed because they are more likely to keep going when things get challenging. I value honesty; I value truth seeking; I value commitment and delivery. It is not easy to do the hard things but it’s impossible to do the things you don’t set out to attempt. I’m awe struck at the many amazing things that Amazon has built.

What defines success in a career? Certainly one yard stick used is promotion : At Amazon being promoted at higher levels is both recognition of career growth and is a matter of timing : opportunity and luck. You need to be in a position long enough that there is opportunity for you and your team to demonstrate completing career promoteable results, and you have to be capable AND you and your boss and those around you (feedback providers) have to be motivated to document and affirm both results and capabilities. Sometimes you’re not capable, sometimes you’re not motivated, sometimes your boss is not, sometimes circumstances don’t line up. And sometimes you choose not to. My own career trajectory involved different mixes of all of those. Don’t ignore your career, but people can’t feel that the only thing motivating you is career progression.

In my last year - I would at times feel a surreal sense of being able to see the end - floating above it all - another month ticked by and less than 12 months remained and now days. How could that be after spending an entire lifetime of working with no specific required date as the end? And I have had some worries about whether I will matter without work. Will anyone need me and my time and my attention each and every day in the same way that Amazon has “needed me” during my time; especially as a Director. I have given my ALL to Amazon for 8 hours+ a day for a long time - but I’ve never given it everything - that is too dangerous - I don’t live to work - I work to live. I work hard; I try to work empathetically and smart and help others - but I am ready to let others carry on. 😉

The graveyards are full of people who thought they were irreplaceable. Just like 1000s of those before me I’ll be gone and you’ll carry on. Keep going!

Closing with the Old Fart statistics

I am employee #1,037
I am older (more tenure) than 99.944% of the company

Cheers
mbg

Subscribe now

I’ve been in the software industry for 30 years. In many ways that makes me an industry dinosaur… and those all died out. So I guess you’ll have to figure out if I am wise or shortly going extinct. The ”I” word that I wanted to write about is not the I in AI, and not the I in Innovative but instead is about : Iteration. In fact I’d posit that most innovation is the result of constructively applied consistently improving iteration. Most people encounter innovations at the break through moments- when they amazing innovation appears with fantastic impressive results - but those innovations are most often the result of countless iterations.

My observation is that one common element in the recipe for failure in building software (not the only one) - is when teams fail to iterate and instead go for the Big Bang, the brand new shiny complicated solution - instead of the iterative release. The alternative is to build the solution that iteratively layers on prior learnings. What is the smallest change necessary to release the next viable feature to create value and to be in production. From my earliest days at WebLogic - we built software daily, tested nightly and we shipped often. (The speed of releases in those days of Enterprise software was glacial compared to today)

When I or my teams have attempted or I have seen other teams attempt the big leap - that requires years of effort and is released in one big push without iterative releases without built in feedback - they have largely failed.

Looking back at big milestone innovations - we often miss seeing the iteration behind them. Ascribing their existence from ex nihilo. Lets look at a few examples :

The lightbulb : Edison and his team conducted thousands of tests on many different materials. One that focused on bamboo filaments, involved 2,774 attempts. Edison famously said : “I have not failed 10,000 times. I have successfully found 10,000 ways that will not work.”

The Internet : burst onto the scene in April 21, 1993 with the release of the first World Wide Web browser : Mosaic; followed by Netscape Navigator in October of 1994. Which unleashed a .com Internet E-commerce revolution. (Which is still on going). But this all had its roots going all the way back to 1969 when the ARPANET network first sent a message between UCLA and the Stanford; and then lots of iterative programs and products : TCP/IP in 1983, the concepulation of the WWW with HTTP and hypertext linking in 1989 before actualization in 1993 of a browser.

The iPhone : amazing in form and function when released in June of 2007. However this brought together many iterative innovations into a single form. Built upon the release of the iPod in 2001 that handled storage and audio. And the first Personal Digital Assistant (PDA) the Apple Newton that was released in 1993 and at the iPhone release was dominated by the BlackBerry. Combined with cellular technology, chip and radio transmitter minutarization as well as cellular network expansion, the iPhone stood on the shoulders of many giant predecessors.

While here we may as well pause to talk about Apple Vision Pro : did it fail because of price point, because of features, because of infrastructure? Maybe all of them, probably worthy of a much more focused deep dive -but an interesting juxtaposition to the success of the iPhone.

And lastly the most recent innovation : AI - Artificial Intelligence. ChatGPT in 2022 : seemed like a innovation that came out of no-where. However ChatGPT has foundations that depend on all the machine learning developments, on the development of neural networks and large data processing that identify patterns and “learn” as they are exposed to more data. And other developments in natural language processing applied to large data sets : Large Language models. These LLMs have billions of parameters and become the backers for chat bots that in response to questions (prompts) can generate statistically predicted text results. They are artificial because they were made by man - and labeled intelligent because their output appears/reads/sounds very intelligent.

And each of these innovation products came from many many iterations and as you look at their adoption and impact they in turn also relied on broad infrastructure contexts within which to operate. The lightbulb was only as useful and impactful as the capacity to generate electricity - and the electrical infrastructure (wires) to connect to people's homes. The Internet was as only as connected and as useful as the power of the computers connected, and the infrastructure connecting them - which ranged initially from physical wires, to utilizing the ubiquitous phone lines and now use high speed of fiber optic lines or the wireless infrastructure that the entire cellular network relies on from the first brick phones of the 1980s to the smart phones of today that started with the iPhone. And finally AI is reliant on massive compute capabilities - which is why Nvidia stock has skyrocketed because their GPU chips offer parallel processing power that is required by computational intensive tasks required for training and running LLMs.

So when my teams come to me with bright shiny new ideas - I always try to ask - how can we release iteratively. Including how can we get feedback on designs and conceptual prototypes even before we actually write any software. How can we break down our full North Star product vision into phases - release software early and often and keep iterating.

Here is a small example of something that I worked on at Amazon that is now comparatively very very big. Fulfillment By Amazon has shipped 80,000,000,000 units since 2006. When we launched FBA in April of 2006, here are all the things that FBA Sellers COULD NOT do :

• Sellers could sign up for FBA but it was not an automated registration system. The registration workflow kicked off a series of 25+ tickets that teams had to manually configure to enable a seller for FBA.

• Sellers couldn’t see any FBA orders in Seller Central - in either a UI form or in a report.

• Sellers couldn’t request removal of your inventory from FBA - except by calling or emailing our operations specialist : Chad Goelzer

• Sellers couldn’t see in Seller Central - the quantity or state of their FBA inventory

• Sellers did not automatically get paid if Amazon lost or damaged their inventory

• Our systems did not have the ability to charge any FBA fees - we had a public rate card - but at launch we did not charge (nor could we have) any fees

But what Sellers could do was create an inbound shipment, have that inventory received at an Amazon warehouse. Sellers could create offers that were designated as FBA, and when inventory was received those offers because buyable on the website, customers could buy those offers on website and Amazon would pick, pack and ship it. We prioritized launching the key features that enabled us to test with sellers and learn. And then we prioritized releasing feature after feature to expand function and entitlement - something that the FBA team is still doing today. And all of that iterative innovation of FBA itself relied upon the innovation that Amazon had undertaken to build incredibly sophisticated operations and fulfillment capabilities. FBA would not exist if Amazon had not first built a great fulfillment network to open up to Sellers. In 2005 we launched with 5 sellers. In 2006 - we ruthlessly prioritized the critical features to launch and got adoption, got feedback and kept going. In 2024 - there were over 55,000 independent sellers.

If you want to innovate - you have to iterate.