Artisanal Concatenations of Sentient Emanations

I have had a curious thirty days.

You know that old wedding saying, “something old, something new, something borrowed, something that scales three orders of magnitude?” I mean, that is basically how it goes, right? What I am seeing out there is kind of like that.

I have had two meetings within the space of one month of each other that have demonstrated technology advances for LLMs that will unlock multiple orders of magnitude of context window scalability. One of them was with a complete stranger, and one of them was with a friend of almost thirty years.

In the case of the friend, this was one of the people who showed me the BREW SDK from Qualcomm and sent me down the path of being one of the first ten mobile developers in North America to ship games to carriers.

I bring that up because I made a post on LinkedIn about the way that mobile adoption happened. It started slowly, and then one day it was everywhere. This was a journey of nearly a decade, and it was constrained by Moore’s Law, among other things.

But what happens when you don’t need to conform to something like Moore’s Law? It was one of the reasons that mobile phone adoption followed a slowly rising curve over eight years. 

Instead, we have concurrent waves of innovation happening where the gains will not be slowly rolled out in n log n improvements per unit time, but one day, boom, your LLM vendor drops trou and there is a thousand times as much junk as there was before.

What this led me to do is come up with a corollary to Moore’s Law. The only right thing to do is call it Lesse’s Law:

“Sometimes technology advances will leap forward unbounded, resulting in hyper-incremental shifts in habits and livelihoods, in ways we cannot predict.”

I was recently in a roomful of people trying to explain the pace of advance we are about to see. It is like we went from pushing a slab of dirt covered with dried beans to feed the people in the cave to hauling tonnes of beans across the country on a maglev train to feed an entire city. Almost no other metaphor comes close to describing the leap forward we are about to take.

While these tools do not solve 100% of all the problems we experience in our day-to-day lives or jobs, they will come up with really good improvements and optimizations for much of our work, up to eighty percent, according to one of the studies on what kind of work LLM technology is good at.

I wish I still had the link to that study today.

This is not some kind of cheering. This is not some kind of Zoomer dooming either.

This is just a warning that it might be worth wearing your seat belt for the next couple of years.

Things are turning out wildly different from how I thought they would just three short years ago. And that will continue to change over the next fifteen hundred days.

Thank you for fastening your safety belt.

I am humbled by the reply to last week’s blog post. Thank you to everyone who responded publicly and privately to me. Writing that post was an ordeal. Do you know how you have a little angel on one shoulder and a little devil on the other one? I could feel my little devil guy picking up and putting down my fingers on the keys as I was typing my GDC write-up. When I was done, it felt like I needed to lie down in my bed, staring at the ceiling and slowly smoke a cigarette, like in the movies, or the nineties. It was good for me.

By Friday, the humility was gone. It was replaced with full-blown panic. How the hell do I follow up to that post? I drank the whole potion. I ate all the magic beans. Everyone expecting the rainbow-belching unicorn to leap from my fingertips is going to be disappointed, perhaps angry when they realize there will be no cash money refunds.

That being said, I have never had so many people laughing at my “old man yells at cloud” schtick. This is a better joke than my track record as Amazon’s worst Affiliate Marketer. That one is a shout-out to my OG readers. IYKYK. I will stop narrating through the fourth wall now. It is weird and uncomfortable, like when you are alone in a movie theater and the only other person to enter the theater comes and sits right next to you.

Without further distraction, let’s get to work, shall we?

Content people are at war. If you have any friends in entertainment or friends who like entertainment, they are all busy selling hats, saying “Make Content Great Again.” They hate the AI, even though AI is right in the middle of the word entertainment. ENTERT-AI-NMENT, in case you missed it.

I am equal parts transgressor and peacekeeper in this war. I bought the hat and scribbled “You are absolutely right!” underneath it and laughed far too long at my own joke. Maybe peacekeeper is too strong, but I do sit down and watch this all unfold while occasionally taking a free shot at someone busily drawing a picture in the middle of the warzone, mumbling “I am making art.” I need the experience points, you know, whether they were fighting back or not. I still have to level up at the end of the day.

So the war on AI is happening. People are canceling each other, and their projects, and rescinding awards, and probably getting uninvited from Christmas parties. If you love this hacky AI bullshit, you must be a monster. Sure, you kiss your kids good night at the end of the day, but that’s fake. If your daughter knew your opinions on AI, she would scream at you until you went away until her new parents arrived with ethically sourced artisanal opinions.

That is not how it actually works, so don’t get excited.

I spend as much time battling people in the digital streets of the internet as I do watching the ebb and flow of opinions and ideas, as each group of deeply entrenched borderline sociopaths tries to convert their opponents. This week, I observed a new tactic, and I am here to develop my mouthfeel for it before I bring it out onto the internet. And by the internet, I do not mean Reddit. All the bosses in Reddit have red skulls instead of numbers on their nameplates. I have to generate about sixteen metric tonnes more LinkedIn tears before I have enough levels to go in there and avoid getting one-shotted.

This week, I have observed the rise of the human slop counterargument. I lay no claim to this phenomenon. I am like that recently deceased ape-observer Jane Goodall, watching internet dwellers in the mist and documenting it for science purposes.

The argument here is one of those Aikido moves, where you take the energy of the attacker and use it against them. Yes, I had to Google that. The idea here is that you can essentially dismiss the people who point at everything and go “slop, slop, slop” like some kind of broken record player as being incapable of making good content… Whether it is books, movies, games, or anything creative. They are merely lashing out at the world because, on their best day, their outputs are mediocre and would not survive the gauntlet of customer review. To put it bluntly, almost everything is already slop, and there are very few things that are quality creations of any kind, anywhere.

I buy into this. We had the tulip-bulb crisis way before we had Italian brainrot. And while Italian brainrot is one of the S-tier slop productions, it is now being heralded as the new punk rock.

Half to three-quarters of you are gone by now. Some of you have come by to see what the second act is to “angry man screams about GDC,” and this is not as good as the first act. We wandered into the heart of AIAIO! country, and you didn’t spend your money on that.

For anyone who wants off the conveyor belt, just extend your hand. Privately tell me which group of people you feel most safe with, and I will point you in their general direction.

I am going to pause for one sentence to let some people get redirected to the comforting sands where they can stick their heads.

For those of us who enjoy the violence of ideas in the pursuit of glory, I am glad you are still here. And that is what this is.

The threat of AI slop is not in some diminished quality of experience. We already have games getting shut down three weeks after launch. Many books don’t make the top ten list (all but ten? Hello math?). Movies are shown to empty theaters and get shuffled off to on-demand, while still watchable on the big screen. I call this “Tuesday.”

What I don’t understand is that if this already happens, what is everyone’s fucking problem? There was a coffee cup in a Game of Thrones set long before some celebrity shot you all three of their middle fingers in contempt without SAG’s permission. Is it the sheer volume of what is coming that has you disturbed, or the fact that your own human-generated content will only aspire to fifty percent on the slop-o-meter and get buried by someone who can prompt more than four times an hour?

These are tough questions, and I am not sorry.

If you can sit down and figure out exactly what it is that is bothering you, it will help you move past it. There is stuff happening in the world today that is very much “coming in 2027” in my head. GDC showed me a silent minority of LLM caretakers who are approaching other galaxies at near lightspeed. Some of them are even doing this from the comfort of their car, with voice-assisted tools.

I am not expecting to convince anyone of anything with this rant. Half of this is a self-aware overreaction to my fifteen minutes of fame. What I am hoping for is that one Grumpy The Dwarf on the subject of AI decides to soften to Skeptic The… uh… Smurf… long enough and far enough that they stand something up in production that catches fire. I do not literally mean catches fire, that would be bad. I mean, it gets its moment in the sun and generates a pile of money and fame.

I feel good believing that I helped nudge them in that direction just a smidge, to see that everything human-made was already on a spectrum of slop, and so why not just go ahead and ship something with AI tools anyway? It will only get you to where you were already going faster.

If I can do that once, for one person, then I can rest satisfied knowing that I have done my part.

Everyone knows I have a love-hate relationship with GDC. If you do an ASMR whisper of GDC in my ears without a trigger warning, I will start crying and laughing hysterically at the same time. It is a complex emotional thing because GDC got me to where I am today. It also went out to get milk in 2009, and never really came back to me.

If you are looking for a coherent, balanced view of the show from a “yes and” for all parties involved, run the fuck away. This post is half warning, and half of a form asking for a DNR signature by a loved one.

How you can issue a press release where you lose ten thousand attendees out of thirty thousand attendees and attempt to say it was a solid fight and there were good people on both sides is beyond me. This is not what happened. GDC just got handed their own ass by the world, and after each disgusting bite, they are claiming it tastes delicious with hot tears streaming down their face between each painful bite.

Never has it been clearer to me that gaming is not what everyone thinks it is. We thought it was this massive community of players, publishers, marketers, and operators all joining hands and singing together to bring swipes and clicks and joy to everyone.

Instead, it is a group of people who got halfway to the heavens, building the Tower of Babel, and God just snapped his fingers, and we are no longer speaking the same language.

The only thing left to do is for all of the tribes to wander back to our mud huts, break out the smelliest hair-bag of fermented mare’s milk, and wonder what the hell went wrong.

I am not at all sorry that this is not what you expected to read, but I have let enough time pass that my take on this subject is clear.

GDC is dying.

Whether this is some phoenix bullshit, or if it’s just gam-gams running out into the street with just her knickers on after secretly not taking her medication, is all that remains to discuss.

Let’s break it down.

Here are the tribes that are politely pretending they like their relatives at Thanksgiving, and promising to behave:

The Roblox Renaissance

MobileMaxxers

WishListMeOnSteam

AAAbsolutelyInDenial

AIAIO!

I am going to have to describe each of these groups of people, even though you all should have Supreme Court Level Perception here: You know it when you see it.

The Roblox Renaissance is clear. The kids are back in the driver’s seat. This is the real meritocracy of making games now. BrainRot and Also-ran-Farmville are king. To people who “make real games,” the kids these days have invented their own language, and while everyone else wishes someone would call security, they are giggling and bouncing around the room, going “6-7! 6-7!” and laughing like it is the best thing ever. Also, if you are not fucking terrified of this already, they have their own happy meal toys more often than Mario and his Lizard-horse-tongue-fetish-mount.

MobileMaxxers are here, fighting for the last joule of energy at the heat death of the universe. They will cut you as soon as look at you. Getting squeezed between Ads and ARPU, they have seen the birth, life, and soon-to-be end of a whole ecosystem. We are Rutger Hauer, waiting to release the pigeons and fade to black. Everything is eating everything else, and in the end, only the hivemind will remain — a cacophony of angry birds meets clash of clans meets merge-two mechanics falling down a big hole in the world. If you did not understand any of that, you didn’t click enough ads, and you probably are going to play the Monopoly Go Co-op Race Game (best feature!) despite what Reddit cardmaxxers KNOW.

WishListMeOnSteam. I cannot even write this without weeping and shaking. One in three people is out of a job in the whole games industry, and they all believe that they can win the Hunger Games. Everyone is excited to be Tribute, and of the millions of games about to come out, everyone is grasping Gabe Newell’s ankles and keeping anchored to the world of Pixels-by-humans by the sheer force of will of “AI is worse than cigarettes” warnings he is slapping onto AI games on Steam. Everyone has piled into this ramshackle shed without realizing that there are six comfortable chairs and eighteen billion software developers. The music has stopped, and everyone is running to sit down. If you don’t have Jimmy McMarketing in your corner, you will get trampled in the rush to sit down and not even understand why. The worst part? Most of these trampled-on designers have great games that will never earn enough to pay back the cost of development, let alone the submission fee.

AAAbsolutelyInDenial. I am tempted to just scream ARCOMAGE and then leave you all wondering if you need to call the police or a therapist. What happens when sixteen MBA’s go through a revolving door between all of their jobs every five years and mostly spend their time day drinking under their desk? Go get a job at a publisher and find out. I am waiting for Soccer Players to become skins in the next Shooty McWar game, and I am hoping that the John Madden estate keeps his 3d model out of Fortnite on general principle. It will eventually come to that, and I am not afraid to be the first predictor. AAA is in a world of self-inflicted hurt, and sequels and remakes, and remakes of sequels will only take them so far in a world where public quarterly reporting is needed. EA at least decided to get off of that merry-go-round, and the only reason you decide to do that is you know what is about to happen to your company’s share price, and the answer is not “get another Bentley to park at the east coast summer house.”

AIAIO! is fast becoming my tribe. Automate the shit out of everything. The solopreneur is back, baby, and the sooner I can start grunting out my top three ideas that I cannot afford to pay people to build, the better. Right now, you will never find a better time to become a Will Wright, or a David Jaffe, or a Dani Bunten, or a Jon Van Caneghem. We are doing our own thing because all of our former tribes-mates hate us for the raw admission of our thirst for this moment.

If all of this sounds unnecessarily dark, it is because it is. We are living in dark times. I have jumped from tribe to tribe in games and have enjoyed the early conversations and the rise of a community and the camaraderie over the years of knowing “hey, we all make games.”

But this year felt different. When speaking with people, it was like visiting a country where you had a dog-eared translation book, and you had to slowly look up the words and point at them to each other because you didn’t know how to pronounce them.

Your brother from twenty years ago? You avoided his gaze. You don’t speak anymore, because he has a t-shirt that says “You are absolutely right! I do love AI tools!” and Mother and Father uninvited him from Christmas and Thanksgiving.

The whole thing is rocking and creaking and straining under its own weight, and it is about to collapse.

I commend the company behind GDC for publishing the stats of the show, but 30,000 minus 10,000 does not equal “thrilled.” Not even close. You earned your “Be Brave” conference scout sticker.

I get that there is military action in some parts of the world. It actually affected my GDC badge, if you can believe that. I get that some people did not come due to current administration reasons.

The size of the impact also speaks to economic shifts. I met more students and juniors at the show than ever before, and that is a good thing, but they had lots of questions. And not a lot of people had good answers.

Some people did have answers, though. I have had my potential GDC talks denied, with respectable enough reasons, revealed to me by honest and hardworking advisors, that I chose to “spite-educate” people at the show. GDC’s refusal to accept my talks did not prevent me from being one of the 4 most useful people in a room of 80 contractors and freelancers, all of whom had questions on how to succeed in today’s environment. I felt good about helping out and giving some useful information alongside 3 other generous donors of their time.

Getting back to my rant. The show has changed. We are no longer Pangea, the Earth-Continent-As-One. Maybe we should just acknowledge that. The sessions were packed with people searching for knowledge, and the young people were awesome and hopeful between moments of fear.

The show hotel lobbies were no longer elbow-to-elbow. Every time I wanted to go for a business meeting at any hotel in any direction, I could sit, hear, and talk without screaming.

The show floor? A joke. The exhibits were tiny, and instead of wrapping around to the north hall, there was “some of south hall” filled, plus three Ringling Brothers stages for the event, placed strategically to prevent the “where are all the vendors” fear sweat from showing.

All of this at the price of a week of hotel stays in San Francisco.

That is not sustainable, my friends.

There is a future change coming to GDC, after a year of radical change.

The show will probably double down on education and attendees. Surveying this young and talented crowd will make the magic eight-ball shake result in “move to a cheaper city.”

Atlanta or Chicago, take your pick.

Business Development will belong to the people of DICE. More than one person said they are only going there next year, not GDC.

The show? The floor? The exhibits? We roll the bones…

And the bones tell us… Nothing.

I don’t love that when I post about AI ideas online, I get called stupid or insane or unethical. I don’t love that everyone has dogpiled into the sea of wishlists and believes that everyone there will be all right. I don’t love that mobile consolidation is accelerating.

But all of these things are happening.

I always tell people. Today is the best time to contemplate a lifestyle making games. It is also the worst time to contemplate a lifestyle making games.

Every year that keeps getting more and more true and it is getting stretched in a line that gets thinner and thinner.

And I do not know when it will eventually snap.

See you next week.

It has been a long time since I have come on here and crapped on someone who did something bad— possibly weeks! I was having a “something something ai” conversation with a good friend of mine recently. Yes, the very same gentleman with whom I got into a heated online fisticuffs match over Unity download fees a few years ago. We exchanged “How Dare You?”s, and “Good Day Sir!”s. We went back to being friends again. The power of the internet, you know?

Does anyone remember Midjourney? And I don’t mean “mid” Journey, like the band’s first album, but Midjourney, the image-generating tool from a few years ago. It started off its business doing something novel… They put the UI for their product into Discord. Oh, now you remember. It was horrible as a user experience.

Or was it?

A lot of the game-playing community is on the edge of their seat right now because an Activision C-level person from the Bobby Kotick and Brimstone Era of Activision is now CEO over there. I actually met Humam at one point, and I thought he was a nice and smart individual. If anyone is going to stab me with a red, burning hot pitchfork and toss my dying carcass into the pits of hell, I will think “well, the nice man is just doing his job” before I start screaming in agony and falling for all eternity.

So what is everyone nervous about? Something something IPO. Something something advertising. It is honestly a refreshing change from people clutching their pearls over AI. It is nice to see a different level of outrage for once. Talking to people who hate AI is like playing one of those coin-operated whack-a-mole games, but it is almost out of batteries, so the little moles are moving REALLY SLOW.

Let’s get back to IRC. I mean ICQ. I mean Discord. You might not have seen what I just did. While Discord, which started as a second company from someone who loves to spell MMO, is preparing for a community-based IPO, everyone here has seen this pattern before. The community shows up, the community gets happy, the community gets monetized, and the community moves on.

We are in the latter half of that conversation. After the investors in Discord git payed, the company will join the public markets and join the mad rush of preparing quarterly income statements every year. “Number goes up” is the only thing that matters to the new Discord owners, and they will start to iterate on product ideas. Eventually, they will put a check box beside every possible revenue-generating feature except “Audience Eyeballs” except perhaps “Audience internal organs.”

Just because we haven’t seen the last one implemented does not mean a public company CEO hasn’t thought about it. I just got done telling you that even the nice C-level people I met do Daddy D’s work from downstairs from time to time, and they are still home in time for dinner.

Now that I have set the stage, I want to talk a little more about Midjourney and Discord again.

I think that Midjourney’s launch on Discord was Discord’s biggest failure.

If you remember attempting to get an image out of Midjourney back in the day, it was like trying to take your clean underwear out of the laundry during the spin cycle. Sometimes you get a sock, sometimes you get someone else’s sweater. The process was fast-moving, filled with other people’s stuff, and very painful. It was almost at an MVP level.

Almost.

If you go ask the internet how many people suffered this atrocious experience, as one does, the internet will give you this mind-blowing stat:

At its peak on Discord, the Midjourney server reached over 20.9 million total members by September 2024, making it the largest server on the platform. Daily active users fluctuated between 1.2 million and 2.5 million, with over 1 million concurrent users online at peak times

That is the size of Canada, and way more people than that started using NCSA Mosaic when fax machines still ruled the earth.

I don’t know how many product managers and engineers it would take to have made that a good experience, but if I was the CEO of Discord, I would have went full Gary Oldman on that.

With twenty million interested users, Discord at that moment could have transformed the front door of the internet and taken a shot at replacing browsing the internet with Discording the internet.

“But wait,” you say. “That would be expensive,” you say. Uh-huh, sure. And there would be a line out the door longer than a Taylor Swift concert inside an Apple iPhone launch of VCs from Sand Hill Road, ready to drop nine figures of cash on it.

This was a catastrophically huge strategic failure on the part of Discord. The resulting slap fight between “the community” and “Q4 profits will be micenuts without Walter Goggin’s Goggle Glasses ads for GoDaddy” is essentially well earned. (https://www.youtube.com/watch?v=K2Gj9BufAqo)

That might sound loony tunes to you, but a few things:

First, that Walter Goggin’s guy’s face is all over my YouTube music. (No, I don’t pay for it. Yes, I think I am getting better music than my Spotify free account.) I love him as an actor, but it’s wearing me down slowly.

Second, my World of Warcraft guild (recruiting ranged dps! Contact me for details!) has already started conversations about where we will migrate to after the IPO if Humam decides to drag the excretions of an ad sales force across our eyeballs for the sake of their quarterly financial reporting.

I don’t think it is an if. I think it is a when.

Considering that they already didn’t take their best shot to win all the marbles?

I hardly care.

A long time ago, I discovered the Serenity Prayer. For the click-impaired, that is the whole “Strength. Patience. Wisdom.” thing. At the very beginning of my career, I loved to use Strength. I used it for everything, in fact. Patience? Wisdom? Those are choices for sissies. Good old strength!

I used to smash through walls at work with such ease and consistency that it terrified people. There was no problem that Strength could not solve! This remained true for about six years.

Eventually, you do have to try the Patience thing or the Wisdom thing. They are equally effective tools for solving problems, especially when you start to transition out of pure IC roles.

Most of the best C-level people I know are masters of deploying Patience. “Do not make a decision until you absolutely have to” is something I have heard enough times that I just had to try it. It takes a while to get accustomed to it, but now that I have a good mouthfeel for it, I recommend it.

One of the biggest flavors of Patience I have had to learn was Patience with other people. I was very quick to throw people into two distinct buckets that I struggled with: Stupid, or New. This was very easy to do, and forgive me, dear reader, I did this more than was necessary.  I have said my ten Hail Maries and a couple of Rosaries accordingly.

Now that I earned the “Haz lurnt Patience” cheevo, I am proud to say that most places I start, I cheerfully tell people that I would like some help because I am either Stupid or New. This tactic helps you get help for three weeks. After three weeks, you are clearly no longer New, so in the fourth week of working with a new team, I recommend that you scratch this self-deprecating means of getting assistance from your playbook.

Thank you for getting this far. I am almost to where I need to go.

I now give everyone a lot more faith in being New. Especially when they are 100% ethically-sourced artisanally new.

In fact, this has become something I have started to take some pride in. I am very focused on helping people who are New.

If you spend any time sifting through the rubble of my LinkedIn posts, you will see I have four or five years of largely self-therapeutic blog entries. I also have many links to short clips from podcasts and links to presentations and classes, and strong opinions, which are often weakly held.

Over the years, I have taken on dozens of people as mentees. I have been more explicit with some about how much I am mentoring them, and sometimes money changes hands. I do have six kids; I cannot mentor everyone for free as much as I would like to.

This is why I decided to fork over some money to the good folks over at Thinkific and start posting some material for first-time engineer managers.

I am still available part-time as a mentor. I have pricing for people with L&D budgets, as well as pricing for when it is coming out of your own pocket. It is not quite scalable, and I often mentor people for four to twelve months, and then they accomplish whatever goal they were intending, or we have concluded that spending more money on that is throwing good money after bad. I am always happy to spend forty-five minutes with you if this is interesting to you or someone you know.

At the same time, I am also posting some of the parts of the mentoring and coaching that I do at the aforementioned link. I am breaking down all of the things you need to know to be an effective engineering manager, at least by my own measure as an off-again-on-again VP of Engineering.

You can find two samples of that material here:

The first is an overview of the ten-week conversation I have with everyone that I wish to promote to a manager on my own team(s).

The second is the first of the topics covered in the syllabus above, a conversation on some of the bigger challenges you will face in your new role. It was not a lengthy conversation, so I decided to put it out there for free.

In the future, some of these will include a price tag. I am still trying to figure out what makes sense there, because all of the online platforms out there cost not-zero dollars.

In the meantime, now you know about my Leadership Lighthouse, designed to steer you clear of the rocks in your engineering leadership role.

It will also contain random things that help everyone, like how to understand what hiring managers think about when they see your resume.

I hope some of this is interesting to you, and I hope that by the time I start plugging in some numbers for pricing, you will support this project with your dollars in addition to your eyeballs and kind LinkedIn comments.

Thank you as always for your time, and see you next week.

There are a lot of people who repeat things in engineering that sound fair and true and good.

And some of those things are completely wrong.

Have you ever heard someone say a variant of the classic saying, “Everyone knows it takes a woman nine months to have a baby. But you Americans think if you get nine women pregnant, you can have a baby in a month.”

I mean, Elon Musk sure seems to be trying to test this hypothesis.

The problem with this expression is that it is absolute psychobabble.

Software isn’t babies, and if King Solomon was trying to figure out how to identify who owned a software project by splitting it into a bunch of pieces, everyone would look relieved. It would be the worst “You ARE the father” episode of Maury anyone has ever seen.

If you have no idea what I am talking about, that is fine. Soon, you will be old enough to purchase your own alcohol.

This ridiculous expression is meant to give engineering teams a “get out of deadline free” card. And it works because most people find out they are about to miss their deadline when it is too late to do anything about it.

The actual problem with this problem is that software parts are generally interchangeable, so for some value of pregnant moms, you can have a baby in some number of months other than nine.

The real challenge is in the timing.

If you are four months away from a deadline, and you have twenty person-months worth of work, there is a universe where you can actually ship on time.

If you have three people on your team, you can get more than halfway there on your own. A little sixth-grade arithmetic also lets you evaluate the value of adding two more engineers to the mix. The challenge is that they need time to ramp into the project, and you have to give them at least a month.

So you will lose about a week of productivity from one engineer who is evaluating all of their hello world level ramp tasks, and you will also invest one month into each engineer for them to be successful. That still gives you a net add of around six months. Sixteen and six is twenty-two person-months, which gives you a safer path to a successful launch.

The reason we don’t have that super fast nine moms grunting out a one-month baby is that most people do not want to do the elementary school level math to size the impact of extra resources early enough in the project.

By the time you tell your boss the project is going to be late, and he wants you to add someone to the project, the ramp time window alone is going to eat any net gains of a new team member.

I have explained this very slowly to people, using very small words.

Repeatedly.

The average C-level executive will stop day drinking long enough for you to go and make your pitch for extra team members to finish a product on time at least twice, and the first time will be too far away in the future.

Random leaders generally don’t think tactically that far into the future, unless they are doing the math of acquiring themselves a schwaggy new Bentley ride. They work hard daydrinking until 4:59 PM; They deserve those sick wheels.

Hiring two more engineers will bloat the costs of the team, and if it is that far away from the deadline, then bossman will be at risk of losing their own bonus. Suddenly, that five-year-old Jetta is blocking the view of the Bentley. While it is good for the team and good for the product, you are likely going to get vetoed that far away from the end of the project because the average suit would love to roll around town in a Bentley instead of a washed-up Jetta.

There is a point to all of this.

You are likely to get shot down the first time you ask. You have to try. It is okay to be shot down with this kind of request. Once. We all get one turn inside the barrel. The important thing is to capture this data so the next time it happens, you can point out the consequence. When someone finds out that the responsible party for the last failure is looking out at them from a mirror, you can get people approved and onboarded in time to make a difference.

I will also add that you probably don’t want to try this with eighteen person-months of work still remaining and trying to hire 9 people to each do one part of the project. Eating nine months of ramp time is hard. You will likely be able to get away with two or possibly three headcount to solve this problem.

So go ahead and get three people pregnant. In four months, you will have a wonderful baby. Maybe you can give it one of those stupid Elon Musk names.

I gotta go now, HR is on their way after reading this post.

See you next week.

This article was almost called “narc-cissists.”

I have inherited multiple bosses I did not want in my career. Two stand out because they were either bad people, bad at what they were supposed to do, or they did bad things to me. I confess to having bad-blindness on this issue. I frequently second-guess how many “bad” checkboxes they check above.

In one case, it was through corporate activity where one team got eaten by another team in a business-related valuation assessment we affectionately call “thunderdoming.” Two VPs enter, one VP leaves. The good news is “something something efficiency.” The bad news is that I suddenly got a new boss I did not want. They made it clear that most of the intermediate leadership in our organization were employees that they did not want.

The boss I did not want was very vocal about us being employees he did not want. That is a really crummy place to be, honestly. Early in my career, I would have said colorful words and done a double birdflip while moonwalking out of the building with two weeks’ notice, enjoying the plumes of toxic smoke that were visible from space emanating from the bridge I just burned to the ground.

However, John became wiser over time, and he decided that he would do his best to help transition his team and the faithful worker bees (who loved their jobs) into the new organization, and just wait to see what leadership wanted to do to solve the “I don’t want you, you don’t want me” organizational symmetry that existed.

This nearly year-long process saw several people declare “shiny and chrome!” and make a run for glory and eternity on their fabulous exits, and some others who went on “hiatus.” Hiatus meant “I am quietly resigning and no, I am not coming back.”

Some of us gripped the broom handle harder and just mopped more fiercely while avoiding eye contact.

This went on for most of a year, with a group of four or five of us huddling together in our own little self-therapy group.

New leadership decided to try to play some headgames and run us through a reality TV-style experience where they would turn us on each other. Some of the attempts at this were more successful than others. There is nothing like being singled out for negative attention and being told to dance, dance, monkey dance, while people throw burning spears at your feet. The idea of this activity was to flush out my boss and get him to make a heroic run to rescue me. The only thing that guy could do was nothing, honestly, and while I shrieked from the center of the clearing, the real prey of the activity remained hidden. I honestly didn’t realize it at the time, and I was furious that my boss didn’t leap in front of those spears, Gandalf-style, to save me. About three months later, I understood it was all part of the dance.

This kind of thing happens, believe it or not. We were being encouraged to do something, and that something was effectively “fuck right off.”

I lasted about a year doing this. One of the fascinating things I learned about this was about the right way to handle crucial conversations, and I want to share that story because the other party involved clearly did not have the benefit of those learnings, because they were a walking, talking counterexample.

Let’s set the stage, shall we?

There is a delicate balance between product and engineering organizations. Sometimes the product team are great partners. Other times, they like to think of the engineering team as “my bitches” and treat them accordingly. You can guess where this story is going.

One of the roles of engineering leadership is to protect your teams and make sure they are being fed and watered appropriately — emotionally and financially, mostly, often to the point of infantilization, if they are very good as a team. I do my fair share of this, even if some people on my teams remember me as a “Darth Szeder” manager and boss. That name arose from gossip randomly on more than one occasion in my career, and I know that says I make tough decisions on behalf of the Empire. I confess to being red-green light saber color blind. I will swing the blade mostly for the “vwoom vwoom” noise it makes, rather than any particular ethical rationale behind the motion.

On this random occasion, I decided to push back on my “peer” on the product side, who had an unreasonable ask. “Let’s find our shared understanding,” I thought, and “Maybe the crucial conversation we all took here will be put to work, and we will both laugh about this all later.”

Nope.

I pushed back on an unreasonable product request, and less than one hundred seconds later, my Slack lit up from my boss, Emperor Palpatine, rattling off a litany of messages and racking up a huge red notification number.

Ruh, roh, Raggy!

Yes, that is right. My teammate just dumped my reply on my boss and waited for me to self-strap the ball-gag on and crawl back into his good graces through the slightly-too-small doggy door.

This was a valuable lesson.

First, I learned that I never want to be that kind of filthy coward. I am not going to pull punches, and not just because I paid for this later. I am not sure how you get through multiple levels of career promotions without learning how to be a good partner with your coworkers and teams. Apparently, this is a very real thing.

I did what my boss ordered and swung my red lightsaber around a few times, maybe employed a few force chokes, and made it clear that this is the new normal in our Emperor’s beloved Republic.

I did not think much more about this incident other than to make sure that we frequently finger-fed our local magnate enough yummy berries and fruit by hand at a rate to keep them appeased, and that I am not here to build a rapport with my partners so much as to just read their edicts sent to me by pneumatic tube.

This continued until employee review time.

In the week leading up to my employee review, I thought about all of the things that my boss could drag me for. There were three one-off incidents where someone escalated something to them, and I expected to hear about some of them. Interestingly, when the escalation was made, the problem never recurred, and in two of the cases, there was zero attempt made to work it out with me. It went to polite compliance immediately, and there was never a disagreement or fight.

What did I learn? That list of three things was 100% of my end-of-year review. There were a lot of positives in the product, and the team that I had something to do with, but none of that mattered. The goal was to move me out of the organization, and a fully toxic end-of-year review was delivered with the intent of handing me off to a different division and a boss who is not at the same titular level as my current boss.

I won’t spend some time unpacking the resultant “pay cut” that came with that, due to a lack of re-upping my equity participation with the company, or how one-sided the review was. I asked my then-boss for thirty days’ grace to digest his feedback before giving him a verdict, and that was the last we really spoke about anything until I gave him notice after that month. I have always been graceful about giving my boss the choice of my final day of employment, and he said an additional month was fine. I don’t know that I will forever give this much grace to a bad boss, but it seems like it is the right thing to do.

There is not much of a moral to this story. It is a cautionary tale. I was asked recently about how to deal with some individuals in a struggling team dynamic, and I made it clear that you should not go running to someone else’s boss to solve problems. You risk getting “Szedered” if I may make a term for the end-of-year review that came about. You should try to sort things out with your peers, or else ask your manager how to resolve it. If they want to speak to that person or his manager to help get to a good place, then that is probably the best thing to do.

You do run the risk during these moments of having them do you dirty. Let them. It reflects poorly on them for not having good conflict resolution skills. All I can say is that you should just make sure that your contributions to the business are good enough that when you eventually decide that you want to be up to your ankles in toxicity, instead of up to your neck, you will (a) be able to get that hawt new jerb, and (b) be written down in the Big Blue Book of Quitters as a “regrettable departure.”

Phew! This is a long-winded dark story. And it might have actually happened.

Stay tuned in the future for more awkward stories that involve me pointing to parts of a teddy bear where I was touched during the process of just trying to do my job.

I don’t usually do this, but I am going to defer a controversial post for a week. I have my reasons. Singular reason, actually, and if you need to know, ask Nunya Bidness for an answer.

Vaguebooking aside, I got into a conversation on Slack today that wound up generating the response “Several People Are Typing…”

The conversation was around vibe coding and how to do low-code to no-code backend services, especially for non-engineers.

I am going to give this a name: Derploying.

Right now, I have had the pleasure of working with at least three non-technical people on AI supplemented projects.

“You are destroying the jerbs!”, someone might proclaim, but I disagree. Engineering has borne the cross of “it takes too long to get to Product Market Fit” on behalf of product management due to “the passage of time,” “we don’t have the right good ideas,” and “engineering takes time to build all ideas, even bad ones.”

The ability to use tools instead of teams to find product market fit has exposed this for what it is. A lunatic fallacy designed to shift blame.

So much for this being non-controversial.

It is really important to shorten the window to PMF, and coming up with ways to create derployments is really important.

So what can you do?

Replit mostly solves this problem, provided you do not mind being married to the Replit. What is made in Replit stays in Replit. It is hard to take stuff out of Replit. It does some crazy stuff under the hood.

If you are on AWS, you can set up a specific instance where there is a shared SSH login secret in your GitHub organization, and create a script to generate a series of project directories for you at the start of a project. Put them inside a folder for HTML deployments, and another one for server binaries, and then you will need to reserve some port numbers. That means this can only work 65,000 times before you have any serious problems. If your company is generating that many prototypes, I want to learn more. Please slide into my DMs.

If neither of these things work then it is just as easy to go grab an instance from a company like Hostinger or Digital Ocean, and just use that in the same way as the EC2 instance.

The idea is that it takes about an hour of engineering time to set up the repo, put the right derployment script in a deployment branch, and then you can move on.

If you are spending more than 300 bucks a year on this across all of your prototypes, you are doing something wrong.

If you are inside a company and freaked out by the idea of product managers pushing code to a server, don’t be. This is part of the new normal.

There are going to be two classes of product builders in the future, and by future I mean last year.

There are going to be “prototypers” and “scalers”. There will be blended engineering, product, and design percentages that will be required for each of those roles to be successful, and they will vary from person to person.

The most successful organizations out there will split their teams this way, much like teams were split up from “hunter-gatherers” to “skinner-cleaners” in their focus decades ago.

Everyone is trying to predict the arrival of AGI. From “It will be here next year” to “This will never happen.”

As you might expect, I land somewhere between the two. I think we are about ten years away.

We will not get there on LLMs alone. If you wanted to make some kind of comparison, the easiest thing to do would be to compare AGI to a human brain. After all, AGI means that we will be approximating the output of a human brain sufficiently close to a real brain that it will be virtually indistinguishable.

LLM stands for Large Language Model and represents the ability to infer the next possible symbol that makes sense in a chunk of output, based on what has come before, and what the users ultimately asked about.

I decided to Google what part of the brain an LLM compares itself to, and if you were interested in that kind of thing, you can go do that yourself. For our conversation, let’s say it represents some percentage of our ability to process language. Really excited techbros might say sixty percent. Skeptical naysayers might peg that at twenty percent.

In my opinion, we will hit the maximum possible value coming out of LLM optimization within a few years. I don’t have some kind of law or some math words behind this. I am just trusting my own instinct here after watching four waves of technological evolution since the early nineties.

We are already getting pretty good outputs from these models. Where they fall down is in context management. We are constrained by the size of the processing model relative to people’s power consumption and CPU capacity. How big does your context need to be? And how much juice do you have to pour into the magic box to have it emit the things? That number may be marginally getting better, but as I understand it, we are linearly scaling relative to the supply available of each, at best.

As we converge on the best possible output for LLMs, we will start to see people building simulations of other parts of the human brain. And this is what we need to get to true AGI.

In about a decade, we will have four or five different systems that have the complexity of an LLM. Each of these will represent an abstract version of a part of the human brain. More importantly, we will have a governing system that decides which system or systems are needed at any given moment.

That combined set of systems will be reasonably deep, and I think it will behave functionally identical to a human brain. It will have flight or fight instincts, the ability to be pleased with itself, and a vast sum of both contextual knowledge about just about everything that anyone could know, plus some extra wetware to do different brain functions above and beyond hallucination reduction.

This is awfully handwavy. I think that it is the minimum of what we need to accomplish to generate AGI. We will not get there with just LLMs, even if we have experts writing prompts that make people wonder if their computer has the feels.

That’s it. That is the whole post.

Unrelated to that, Friday’s presentation on Tips and Tricks on Hiring in 2026, also known as “something something ai,” was great. We had a small group of people having a highly intelligent and valuable conversation about things we have all seen in hiring, and I think more than one person took away some new tools and tactics to experiment with when interviewing. I also learned that I should be checking the battery level on my Bluetooth headset. They gave up the ghost halfway through the show, and I had to put my giant air traffic control headset back on. I am going to re-record some of it and put it back into my Leadership Lighthouse site as personal coursework instead of a webinar. I will update you all on how those experiments are going.

See you all next week!

This is going to be short this week, and I do not want it to be, but a brother has gotta git payed. The rent check is due on Monday, you know?

If you have been following along on LinkedIn, you will note I am doing a paid webinar this week. I have spent the past six months furiously coming up with ideas on how to effectively, authentically, and musically interview remote candidates so you don’t have to.

I am not going to wax philosophical here about combating interview fraud. If you want to know more about that, come and listen in on Friday. It will cost you a few bucks, but you know, tariffs. Nothing is really free anymore. You can enter “mofactor” for a 25% off code, though, just to let me know that you read the blog. That is a permanent discount for all future products, not like the limited ones I am handing out here and there.

Real talk, I am the world’s worst Amazon Affiliate Marketer… or you people all have busted mouses. I don’t know which version of the story is worse.

If you are a hiring manager, or you play one on TV, you should easily get your money’s worth on tools and tactics for remote interviewing in the AI age.

Okay, I will shut up about the paid stuff.

I did add some free stuff, too, for job hunters. I added a ten-minute rambly monologue on resumes for your enjoyment. Sure, it is probably a lead magnet. I fully intend to never take profit from people out of work, but my ideas on how to fix your resume are SO GOOD that you will want to attend one of my webinars after you get hired, right? I thought so.

Now that we understand each other, go click the links, please. I asked nicely. If you need some help with your resume, I gave you some free advice. If you are hiring people and want some ideas on how to make that process smoother, I have some “less free” advice.

Thank you for your patience. I have big ideas scheduled for the next few weeks, but this week is all about shamelessly plugging my nascent engineering leadership program. I would love your support.

Thank you all, and especially thanks to Casey Rock, who helped me put together some videos this past week. He did a cute write-up of one of my foaming-at-the-mouth rants about hiring and the kids these days.

You should go read it too.

See you all next week.