He wrote software for controlling hydroelectric dams when he was still in high school. At a time where the tooling was incredibly worse than it is now.
What's your definition of exceptional? Because that fits mine.
Computers, programming languages, and software in general were simple back then. It didn't take much to program them. You couldn't do much with a few KB of memory and a single-core single-thread CPU with basic instructions.
It took some amount of talent to squeeze something useful into computers with those constraints, but the upper limit on complexity also put an upper limit on how exceptional anything could be.
Bill Gates is/was an exceptional businessman. His technical abilities were nothing special.
By your definition, is John Carmack an exceptional programmer? After all, Wolfenstein3d is also simple, if you ignore the challenges of limited hardware and of being an early mover who has to invent lots of things from scratch. Who from that time period would qualify?
Exceptional is a relative term not an absolute one. I dislike Gates but you can't simply argue that he was not technically exceptional because the primitive tools he used capped how complex the software he wrote could be from a modern perspective of software complexity. It is ridiculous. By the same token, any software anyone living can possibly produce pales by the standards of a 2090 dev.
The reasonable take is that Gates was an exceptional dev for his era for the things he did with the things he had.
A defense contractor on a large contract was having problems finishing the project and sought Gates out when he was still in high school. By the time he got to Harvard, he was taking graduate-level courses in computing.
But I'm not going to insist any further because after arguing that it was pretty normal for kids to go around and program controlling software for a hydroelectric dam system, I don't think there's much to reason with you.
I'm not going to get into the debate of ideas vs execution. It's pretty much imponderable. Both are necessary conditions, who cares in what proportion of importance.
I'll just observe that execution remains ridiculously difficult despite AI. Even building the software itself. Then there's all the other stuff, mostly relationships (with investors, customers, employees, lawmakers, society at large) to which AI contributes nearly zero or may even cause damage if used recklessly. The amount of posts I'm seeing from VCs complaining about vibeslop decks!
If you're referring to the famous Henry Ford quote ("If I had asked people what they wanted, they would've said faster horses") I'll first note the obvious: Ford didn't invent the car.
Second, I'll concede that revolutionary innovation happens primarily top-down. Common folk don't know anything about the state of combustion engines or transformer architectures, or how those can be composed into new categories of product.
Third, odds are if you're building software, you're doing nothing of that sort, but rather dealing with infinitely more modest innovation. You would be, in fact, improving a horse, if that were at all possible.
Fourth, when improving the proverbial horse, you'd better have an outstanding sensitivity to the finer problems that customers face. Whether you develop that sensitivity by asking, or sensing in some other way is another question. But asking may not be a bad idea at all.
Fifth, I'll close with an anecdote. The worst seed-stage software startup tire-fire I've seen in the last couple of years was led by a CEO that published the Henry Ford quote, verbatim. A couple years in and they still don't have a paying customer. I'll generalize by saying that I've seen pretty much everyone around me with "the don't ask, envision" attitude fail, while the people who ask or are otherwise exceptionally attuned to the customer problem do much better on average.
Dwarkesh has been skeptical about AI claims for a long time. In his interview with Dario earlier this year, he pressed him hard. So this is a fairly recent change of opinion.
What he says still doesn't map to my daily practice. I use Astra and Fable extensively and they need an incredible amount of guidance and harnessing to produce fairly simple systems. Any decent code is produced with tons of static tooling, agent files, harnessing, and even then, there are myriad mistakes. My main point of skepticism is whether software engineering is verifiable at all, once you factor in architecture, code design, let alone the human element in usability, requirements, etc.
As far as I'm concerned Dwarkesh could be prey of AI psychosis... I don't think a proxy metric like the average daily spend of AI researchers going up proves anything.
The technology has undoubtedly advanced but for me it's still a coin toss on whether we're in an exponential or an s-function.
Taking a step back, it's basically getting very good in one arena, with who knows how much money is being spent refining things. But nobody appears to be working on the other aspects of intelligence that I (as a lifelong sf reader) would expect from an AI, let alone a superintelligent one. There is no permanence of anything except at the "build the next model" level and some notes it keeps and can scan quickly. Instead of an intelligence that focuses on you for a moment, you're getting a new copy every time you interact with it after a certain amount of time. There is no "game loop". It's just there to respond to you and be ordered about. It will never reply in a text chat hours later wondering how your latest meeting went. It doesn't have a concept of concepts. It regurgitates words and code in helpful ways, but there is no understanding there. It doesn't know what it means when it says things, it's all just a (very impressive!) act.
I've been building neural nets in C as a past-time hobby, and I get the feeling that I've been peeking under the curtain too much. The magic is going away. I know they have been getting creative with training, different types of neural nets, and so on. I don't know that I want these organizations working on building an actual sf-like AI, because that might actually human-ending disastrous, but this side road they've gone down isn't going to get there from where I'm standing.
I'm also starting to worry that all I do is regurgitate words and have no real understanding either :)
This is pretty much what we're discussing: inside it it's very hard to tell. You don't even need to expand the timeframes that much for the point to hold.
I understand. My point is that the timeframe of this S curve may in fact be that large. And the curve that steep.
IOW, if an S-curve looks exponential for your entire life, and the life of your children and their children and theirs, does it matter that it's technically an S-curve and not truly exponential? I would argue that no, it doesn't practically matter for the people in that scenario, both options are indistinguishable and change everything completely.
That also matches my observation, particularly in economics once you factor in a scarce resource. But it's not like I have a convincing metamodel to say whether that or the other distribution is more common. But neither do people who fervently believe in the exponential...
Mercedes-Benz turned a net profit of $5.3 billion last year. Even if they had had a bad year, they would have assets - tangible and intangible - and potential for turning things around. There's plenty of stakeholders with, well, stakes in the business and incentives to see it do well.
How does one come to wonder if they are going to "slowly wind down the business"?
"We're stopping investing in the business" is usually synonymous with "growth ran out and we have no idea what to do besides hope and pray and eventually go bankrupt"
That defeats the whole purpose. You have no way of knowing that the version that the provider claims to be serving is the version that they are actually serving
You can always rent a box and serve an open-weight model to yourself remotely. That's probably the only economic solution that halfway makes sense, given hardware prices these days.
Everybody says they pin. Meanwhile, if they can swap in a lower quant and serve more throughput and make more pennies, they do. It's easy to say, "well, KLD is at 5% so come on that's pretty much the same". And then that's the new norm for a bit, and then comes more traffic and they say, well, shit, let's do that again, it's only another 5%...
I have this problem and I was wondering how it would be solved. For now I can cope because I mostly work from home but I’ll keep track of you guys. Best of luck!
What's your definition of exceptional? Because that fits mine.
reply