I worry a lot about fads in engineering management. Any time you proscribe process over outcomes you create performative behavior and bad incentives in any discipline. In my observation, this tends to happen in engineering because senior leaders have no idea how to evaluate EMs in a non-performative way or as a knee-jerk to some broader cultural behavior. I think this is why you see many successful, seasoned EMs become political animals over time.
My suspicion about why this is the case is rooted in the responsibilities engineering shares with product and design at the management level. In an environment where very little unilateral decision making can be made by an EM, it is difficult to know if an outcome is because the EM is doing well or because of the people around them. I could be wrong, but once you look high enough in the org chart to no longer see trios, this problem recedes.
The author really got me thinking about the timeless aspects of the role underlying fads. I have certainly noticed shifts in management practice at companies over my career, but I choose to believe the underlying philosophy is timeless, like the relationship between day to day software engineering and computer science.
I worry about the future of the EM discipline. Every decade or so, it seems like there is a push to eliminate the function altogether, and no one can agree on the skillset. And yet like junior engineers, this should be the function that grows future leadership. I don't understand why there is so much disdain for it.
Process over Outcome is something that I think would be easy for anyone to proscribe to a process that they didn't like.
In my younger years, I was very cavalier about my approach to programming even at a larger company. I didn't particularly want to understand why I had to jump through so many hoops to access a production database to fix a problem or why there were so many steps to deploy to production.
Now that I more experienced, I fully understand all of those guardrails and as a manager my focus is on streamlining those guardrails as much as possible to get maximum benefit with minimum negative impact to the team solving problems.
But this involves a lot of process automation and tooling.
The problem imo tends to be not that there are guard rails in place. It's that they are often build by people that only care about the guard rail part and completely forget that its supposed to be last barrier and that there are other things you can do before you get people to hit a guardrail
What if teams were integrated groups of engineers, designers, and product people, managed by polymaths with at least some skill in all of these areas. In this case, do you think it would be easier to evaluate the team’s (and thus the manager’s) performance and then higher levels of management would care less about processes and management philosophy?
You're describing the GM (general manager) model, sometimes called the single threaded leader. This does work well in large scale organizations...especially ones where teams are built around projects and outcomes but exist for a finite time. Video game development tends to have this model.
I tend to believe in this model because when I've seen it in action, bad GMs are quickly identified and replaced for the betterment of the project.
It can be challenging to implement for a few reasons.
- It is difficult for a GM to performance manage across all disciplines. This model works best when you aren't interested in talent development.
- It's bad for functional consistency. GMs are focused on their own outcomes and can make the "ship your org chart" problem worse. It requires strong functional gatekeepers as a second-order discipline.
That's usually a consequence of bad incentives. Either leadership is selecting for that kind of behavior in managers or they don't know how to properly unselect for it.
If a bunch of crap code gets shipped, it isn't always because the engineers are bad. Often it's because they were given a bad deadline. Same with EMs.
Design moves at the speed of culture; not technology. It took 3 years of people messing with mobile phones before it occurred to someone to implement "pull down to refresh" and much longer for it to be common practice that people just expect from UX. I think people are still learning what they want from an AI experience.
I do think you have to be pretty targeted with your predictions, though. Consumer product design seems to be evolving differently from B2B and at a different pace. Growth curves are different for each.
One of the bigger design battles at a prior company was designers insisting on pull to refresh, and the researchers insisting on removing it due to customer feedback.
The whole thrust of anti-vibe UI sentiments remind me of when Twitter Bootstrap came out. The unlocks were huge because suddenly people who didn't know how to make nice looking UI didn't have to do much more than drop in a stylesheet link and add some classes. Despite that, everyone complained all web sites started looking the same.
And, sure, that was valid. However, eventually everyone started figuring out how to get a unique look out of Bootstrap while still enjoying the benefits. All our modern frontend component frameworks can trace their lineage back to Bootstrap.
We'll see something similar with vibe UIs. Just a matter of time.
I don't see how what you're saying is at odds with the author. At no point did they say Vonnegut was a failure before Slaughthouse Five. Only that he, like many others, didn't produce their opus until later in life. This isn't just limited to writing. There are examples in all fields if you look, both creative and commercial. This idea is definitely at odds with a lot of current SV rhetoric.
Your point, that many people don't produce their magnun opus until later in life, is definitely at odds with a lot of current SV rhetoric. And it's a good point.
If the article had tried to make your point, it would have been a much better article.
Instead, it made a different, much less true point, and had to contort Vonnegut's biography to make it.
"His career looked like a sequence of failures until it suddenly wasn't" is just not true of Vonnegut, not true of Galilei, not true of any of the other "examples in all fields" cites in the article. All of them are people who consistently produced great work from early on, well before their 40s, and then produced a magnum opus that really stood the test of time.
I think the people who need to hear this message are not in their late 40s, but are in their 20s thinking that they have to do it now or they never will. I see a lot of ageism on X from people who are clearly young and inexperienced, but when you actually get into the real world and how things are done you see the vast majority of real success happens later in life once you've been around a couple times.
Experience, on the whole, really does get you further than cleverness, but good luck telling that to the inexperienced.
Something we definitely lose when we age is lack of judgement, and with it the ability to play and experiment. I miss that naïveté and over confidence from my twenties knowing I was allowed to screw up.
Is it really those things, or is it that as we get older we have something to lose? I could go live like I'm 20 again...reduce my spending to nothing, work all day on whatever I believe in...but that would cost me my family's comfort, and I rather like them.
In any field where there is a creative element, progress comes in fits and starts that are difficult to predict in advance. No one can accurately predict when we'll get the cure for cancer, for example, in spite of people working on it.
But that isn't how investors operate. They want to know what they will get in exchange for giving a company a billion dollars. If you're running an AI business, you need to set expectations. How do you do that? Go do the thing you know you can do on a schedule, like standing up a new GPU data center.
I don't think the bitter lesson is misunderstood in quite the way the author describes. I think most are well aware we're approaching the data wall within a couple years. However, if you're not in academia you're not trying to solve that problem; you're trying to get your bag before it happens.
Why do you assume investors don’t know about this? They know some investments follow the power law - very few of them work out but they bring most value.
The very existence of openAI and Anthropic are proof of it happening.
Imagine you were an investor and you know what you know now (creativity can’t be predicted). How would you then invest in companies? Your answer might converge on existing VC strategies.
I don't assume that at all. Investors absolutely know, but investment is predicated on returns. You can't do that if you can't give a timeline for when value will be generated unless your investment is so small it's practically a donation. Obviously you can invest in moonshots, but you don't want to bet your whole portfolio. Why do you think OpenAI had the governing structure it did before it made its breakthroughs but suddenly both them and Anthropic can do insane raises?
It depends on who is creating the definition of evil. Once you have a mechanism like this, it isn't long after that it becomes an ideological battleground. Social media moderation is an example of this. It was inevitable for AI usage, but I think folks were hoping the libertarian ideal would hold on a little longer.
It’s notable that the existence of the watchman problem doesn’t invalidate the necessity of regulation; it’s just a question of how you prevent capture of the regulating authority such that regulation is not abused to prevent competitors from emerging. This isn’t a problem unique to statism; you see the same abuse in nominally free markets that exploit the existence of natural monopolies.
Anti-State libertarians posit that preventing this capture at the state level is either impossible (you can never stop worrying about who will watch the watchmen until you abolish the category of watchmen) or so expensive as to not be worth doing (you can regulate it but doing so ends up with systems that are basically totalitarian insofar as the system cannot tolerate insurrection, factionalism, and in many cases, dissent).
The UK and Canada are the best examples of the latter issue; procedures are basically open (you don’t have to worry about disappearing in either country), but you have a governing authority built on wildly unpopular ideas that the systems rely upon for their justification—they cannot tolerate these ideas being criticized.
Software engineering in general is pretty famous for unironically being disdainful of anything old while simultaneously reinventing the past. This new wave is nothing new in that regard.
I'm not sure that means the people who do this aren't good engineers, though. If someone rediscovers something in practice rather than through learning theory, does that make them bad at something, or simply inexperienced? I think it's one of the strengths of the profession that there isn't a singular path to reach the height of the field.
The web doesn't need gatekeepers the way you don't need a bank account, driver's license, or a credit card. You can do without it, but it sure makes it harder to interact with modern society. The days of the mainstream internet being a libertarian frontier are more or less over. The capitalist internet is firmly in charge.
The real question is whether there is more business opportunity in supporting "unsigned" agents than signed ones. My hope is that the industry rejects this because there's more money to be made in catering to agents than blocking them. This move is mostly to create a moat for legacy business.
Also, if agents do become the de-facto way of browsing the internet, I'm not a fan of more ways of being tracked for ads and more ways for censorship groups to have leverage.
But the author is making a strawman argument over a "steelman" argument against signed agents. The strongest argument I can see is not that we don't need gatekeepers, but that regulation is anti-business.
You’re totally right, and that’s why I think this era will fail.
Web 2.0 failed because eventually people realized to make money they needed to serve ads, and to do that they needed to own the UI. Making it easy to exfiltrate data meant switching cost was low, too. Can’t have that when you’re trying to squeeze value out of users. Look at the evolution of the twitter API over the 2.0 era. That was entirely motivated by Twitter’s desperate need to make money through ads.
Only way we avoid that future is if we figure out new business models, but I doubt that will happen. Ads are too lucrative and people too resistant to pay for things.
ads really aren’t all that lucrative, though, they’re just simple. I worked for a company that was trying to figure out an alternative to ad revenue (we failed) and our people did some research and the average internet user ends up being shown (if I remember correctly) like $60/month of ads, total.
That’s not surprising and is also so depressing. Imagine what the Internet would be like if there were zero ads and zero of the dark patterns that they essentially force every web site operator to engage in. And all for sixty measly bucks a month that we’d spend in microtransactions or subscriptions or whatever. :(
It’s an average, you can’t just charge everyone in the world and have an ad-free worldwide internet. About 2 billion of internet users are probably worth 0.60$ pm while 200m are worth 600$ of ads pm.
Maybe lucrative isn’t the right word but it’s more profitable than the alternative. Call it simple but there’s only so many ways to monetize people that feel entitled to digital works and refuse to open their wallet.
In practice it'll just mean that each MCP provider will have API tokens and it'll be even harder to lock down spending than AWS. Maybe companies will need to have a company wide system prompt to pretty please don't use the expensive APIs too much.
Obviously it’s possible. And just like the early internet ad networks…
You can’t guarantee they’ll be shown and interpreted correct by the downstream LLM, you can’t guarantee attribution later when a user makes a purchase, you can’t collect (as much) data on users for targeting, etc
The biggest ad networks today (Google, Meta) have strong first party data operations, strong first party attribution techniques, and strong targeting, either through intent (search) or profiles (meta).
MCP originated ads really only threaten Google (via intent based ads), and they’re quickly moving I to owning the UX of LLMs, securing their place in this value chain.
My suspicion about why this is the case is rooted in the responsibilities engineering shares with product and design at the management level. In an environment where very little unilateral decision making can be made by an EM, it is difficult to know if an outcome is because the EM is doing well or because of the people around them. I could be wrong, but once you look high enough in the org chart to no longer see trios, this problem recedes.
The author really got me thinking about the timeless aspects of the role underlying fads. I have certainly noticed shifts in management practice at companies over my career, but I choose to believe the underlying philosophy is timeless, like the relationship between day to day software engineering and computer science.
I worry about the future of the EM discipline. Every decade or so, it seems like there is a push to eliminate the function altogether, and no one can agree on the skillset. And yet like junior engineers, this should be the function that grows future leadership. I don't understand why there is so much disdain for it.