The whole LLM development revolution has made me feel like I’m taking crazy pills, and this is one of those observations.
And, as someone who does a decent amount of vibe coding -
Are we really turning engineering into plain text and not a like agreeably defined spec? Is that how we’re writing applications now?
That a list of instructions that may not have an agreed upon definition or compilation process to ensure they’re internally consistent - that’s the foundation of modern change management?
It's not the usage of AI so much as a seeming enthusiasm that this generation of tools are necessarily the right ones.
It's kinda weird because I think this almost loops back on itself, but a fundamental question is like - what is the difference between AI and a library?
What's the difference between typing say "I want to reach out to the google API and get a json blob of my e-mail" and "axios.get(google.com/e-mail)"
What's the difference between "Build me a login screen" and having a good set of react components you can quickly compose into a login screen.
As sort of illustrated in the article. It seems more reliable and potentially less expensive to have a layer that can sort of say "This line here, it's covered by this library/git hook, we'll actually pull it out, replace it with a deterministic layer and maybe put it in a spec that's ignored or verified by LLMs to have the deterministic component, but not interpreted by LLMs"
I don’t entirely disagree with this criticism, but it also feels to me like - progressivism is the flavor of modern bad studio films, but not a cause of bad media.
Like going back to the early 90’s there was this incredible rash of bad films, especially action movies, and that was just the nature of Hollywood incentives.
Because I think the complaint in some sense isn’t the fact that there are progressive movies, it’s the fact that it’s a layer added to all low risk big budget studio films.
But these movies would likely be bad for other identifiable reasons if not for the sprinkling of diversity. The lord of the rings on Amazon isn’t otherwise great.
So sometimes I think this criticism is just feeding into the very thing it’s trying to fight. Industry is great, and it’s great for the reasons described, but I think the moral ambiguity problem as described isn’t caused by diversity, they’re correlated. Bad movies have less moral ambiguity and use diversity in a hamfisted way.
There’s something of a footnote in here that - raises a moral or even spiritual question.
Scott says that one of the great things about superforcasters is that some people who regularly make bad decisions will be able to make better decisions by asking superforcasters questions like “Is this marriage likely to end in divorce.”
And there’s a number of profound questions in here about - what are we all here for?
There’s something of a philosophical horseshoe theory here which is like - I might argue - this sounds like astrology or other metaphysical forms of divination.
And the counter argument would be something like - no it’s completely different because this is true. Which is of course what all the spiritual traditions say about each other.
So it begs the question - is it the solution or the process.
Is using a super forecaster better because it’s “real”; or like I think I could make the argument it’s worse for that reason. Asking your grandmother or spiritual advisor is part of a process of existence. You ask multiple people, you have the experience, maybe you marry the wrong person, you learn and you grow.
> Why can't we have a red-queen's race where non-tech companies are implementing AI, but it's not increasing the total profits of those sectors, just meeting rising customer demands/fighting over the share of existing profits?
But then if this happens - all of the stock market has risen in the promise of AI. If AI eats profits instead of grows them, then the economy shrinks right? So maybe that’s worse? That there is no productivity increase?
> How much do these models need to do before people throw their hands in the air and say, ok this is happening.
It's quite interesting to me that your epistemology is basically "God in a box is inevitable" and not skepticism.
> Now if you have millions of instances running in parallel, all "probabilistic", working on frontier AI research I really don't see the blocker (and believe me I wish I did).
The blocker is lots of things, but for one? Compute. Millions of instances running parallel, all "probabilistic" testing different theories would require million of GPU clusters.
Two? For subjects like Physics - Atoms. You need to be able to run experiments. You need access to the real world.
Then there's the S curve. Algorithmic breakthroughs operate on orders of magnitude. Eventually we'll run out.
I mean he explicitly at the end says there will be a new era of prediction markets:
"This, then, is my prediction for the AI superforecaster future: for basic questions, your off-the-shelf AI chatbot will be able to offer opinionated probabilities superior to those of any human. For more controversial or bias-laden questions, a new era of prediction markets will smooth over differences in brand and model and efficiently aggregate all AIs’ opinions."
But one reason that you would publish it for everyone else to react was exactly what I said, because if you can show people that a superhuman AI believes a certain outcome is more likely, you impact whether or not that certain outcome is more likely. Which feels like a flaw in this version of the future. That in fact you could overcorrect and make the markets less efficient.
I suppose on the one hand, I probably understand what this means?
But I hate when guidance is totally verboten because like -
“Babies should not be shaken at all in the first two years of life”. I think has a _significantly_ different impact meaning then when you say it about screens. I don’t think if my baby happens to glance over at the TV there will be significant impacts to health and development.
And I know that, but when you say it’s just verboten, it doesn’t actually provide any signal, and it’s similarly used for things that are less easily differentiated.
"You can absolutely write bad Pulumi code. You can over-engineer it. You can create abstractions nobody understands. You can make something simple feel clever for no reason.
But that is not a Pulumi problem.
That is a software design problem."
An uncomfortable idea is that these are basically the same thing. There's a reason the phrase "just enough rope to hang yourself" exists. A system that requires more software design <can be> a problem in and of itself.
Right, but if your real assertion is “we have no idea”, it seems you should point your skepticism significantly more towards the people betting $100 billion dollars that self-driving cars are coming next year than the ones who aren’t.
I don’t understand the analogy you’re making, or maybe I think it’s wrong. This is the first time ive ever seen someone say you should outsource your thinking to an LLM, rather than say idea generation.
No one thinks you shouldn’t do 8 digit multiplication with a calculator, But you should understand what it’s doing under thr hood so if you say typo something you can catch when the answer is off by an order of magnitude.
But the same argument applies to AI. If you don’t understand the basics of an argument or the nature of the subject you’re investigating, you can’t tell - not even an if it’s working correctly but if it’s responding to the question you asked. If it applied the right context for your particular situation.
And I think it’s the exact same thing - whether AI is really thinking is irrelevant, students need to understand the nature of how to make arguments and validate information, before they can trust their own usage of AI.
How is EFF asking to police or restrict users use of technology? The only thing they seem to be asking here is for X to continue to generate reports in how they use (or misuse) PII.
The concerns on Grok seem pretty specific: to not take for granted that it doesn’t introduce problems with how Twitter handles user data, not what users can do with it.
From the article:
“These sweeping assurances that corporate restructuring led to a fundamental change in X’s policy and practices around user data should be met with a healthy dose of skepticism, given evidence to the contrary. For example, the company’s quiet rollout integrated its AI model Grok with the platform in 2024, trained (without meaningful consent) on X user data. The company was also subject to a massive data breach in 2025.”
Nothing about groks capabilities or what users are allowed to ask it.
I’ve started blogging about my experience with AI. I’m not special, my workflow isn’t supercharged. I’ve just started writing about a guy(me) interfacing with these tools as if I was leaning any technology like a programming language or a new cloud tool.
And I agree with the authors thesis that the hype is actively harmful. Specifically (and this is a confession not a judgement) if you’re usage of AI is limited to mindlessly vibe coding tools all day long, your missing the actual process of just fumbling through the awkward stage of being consciously ineffective, so you can break through to eventual productivity. And the real productivity probably isn’t as exciting as the game or app you one shotted for fun.
We're a few days out here, but I don't agree that a FAANG company will come out of AI, and I think this is a fundamental error to think there will be.
FAANG is rare/hard. It feels like, we've forgotten there's this whole giant middle of like normal companies. Oracle, Cisco, all the consulting companies. Boring companies with "only" a $500B market cap. I don't think any of the AI companies will synergize the like magic set of ingredients you need to be FAANG, and that's the point, they've grown so fast they "have" to to payback investors.
But it's weird that we call them a failure if they don't hit this utopian ideal that only a few companies in history have ever hit.
The attached video did a good job I think explaining the range of problems, and I could definetly see the value applied specifically -
But like - year or two ago now, Satya Nadella I think outlined this idea as all the things, and I think it has to be carefully applied, but I wonder what this does for collaboration, and like trying to help your grandmother understand a UI you’ve never seen before.
> AI companies have a runway. The question isn't whether they can fly, but whether they have enough runway to get liftoff.
This misses I think what is becoming my thesis is that competition from each other and open source models drive both costs and profits down enough that it ends up becoming a feature of a company with an existing cash flow (google, Microsoft) but is not a business that is sustainable in its own. So no, the AI companies may never fly.
I feel like this gets at something, but we’re not quite there yet at understanding the language to use:
“I shipped code I didn’t understand”
Suggests to me that it didn’t follow the Vercel rule of:
“Would you be comfortable owning a production incident tied to this pull request?”
Why didn’t you understand it, it used a type of loop you have never used before? You read the interface but not the implementation? Or did it really in the end do work that was different from your intentions such that, you didn’t understand if your code met the requirements you set forth? How do you know you didn’t understand it? Did you read it and it was just French to you?
I suspect it’s probably something like b? You’re probably wouldn’t be proud of C (I mean maybe if you’re just a chaos agent) and it sounds like you’ve been coding long enough that it’s not A. But B has been true every time we included a library
This touches in some good points, but I continue to think in this world where everything has changed, more is the same than different, and part of our goals in the language we use is to ensure we emphasize we’re talking about what’s changed, not strictly speaking aim for accuracy.
Ok, so one last question - I think folks like Ed Zitron are too much I. The bear camp, he’s “AI is useless” and I’m not there, I do think AI changes the game. But everyone has been comparing OpenAI and Anthropic to the dotcom bubble, but isn’t -
The historical precedent on being able to capture value of a raw technology is not good. This is why the joke they promised us flying cars and we got ads.
So everyone says “Anthropic is like Uber”. But Uber is a service with people, not an underlying technology with commodity economics.
So in order for OpenAI and Anthropic to succeed (not google they have businesses to subsidize AI). LLMs would have to be the first technology where the businesses can capture value of the underlying technology itself.
So while I would concede that affordability is vulnerable to multiple angles of attack, I think that profitability is as well.
So what makes AI different than all previous technologies?
And, as someone who does a decent amount of vibe coding -
Are we really turning engineering into plain text and not a like agreeably defined spec? Is that how we’re writing applications now?
That a list of instructions that may not have an agreed upon definition or compilation process to ensure they’re internally consistent - that’s the foundation of modern change management?
It's not the usage of AI so much as a seeming enthusiasm that this generation of tools are necessarily the right ones.
It's kinda weird because I think this almost loops back on itself, but a fundamental question is like - what is the difference between AI and a library?
What's the difference between typing say "I want to reach out to the google API and get a json blob of my e-mail" and "axios.get(google.com/e-mail)"
What's the difference between "Build me a login screen" and having a good set of react components you can quickly compose into a login screen.
As sort of illustrated in the article. It seems more reliable and potentially less expensive to have a layer that can sort of say "This line here, it's covered by this library/git hook, we'll actually pull it out, replace it with a deterministic layer and maybe put it in a spec that's ignored or verified by LLMs to have the deterministic component, but not interpreted by LLMs"