Joined 14 years ago
I don't mind the above wording at all, because there's another thing that happens: people seem to have this impression that if you get cancer, follow the recommended treatment, and are in remission... that you're "done", and you "won". It isn't like that. First, the treatments have serious side effects, both short and long term. Second, cancer is never gone from your life, even if it never comes back. That doesn't mean you should not do the recommended and available treatments. Your odds of survival are almost always much much higher if you do, and imo is the right thing to do for the vast majority of people. But go into it with your eyes open. You're past the point where you get to have a conventional happy ending. You'll need to figure out how to be ok with the ending that you can still have. (Which may be a happy one! But not in the way that you imagined when you were younger.)
> Normally, I would say you don't want to be patient zero, but with the attitudes shown by frontier labs, I guess YOLO is the way.
In practice, yes. That's exactly how it's playing out.
I think you left out some part of that sentence, it sort of has two subjects? Companies are not investors.
"As an investor, I only care about a company's duty towards shareholders" maybe?
But no, companies are not obligated -- legally or morally -- to be extractive or abusive in pursuit of shareholder's interests. A company can do whatever it wants, can be formed for whatever reason it wants, and shareholders can vote or divest if they don't like it. The whole "a company's only duty is to its shareholders" is a myth that was widely propagated in the 1980s. There's a kernel of truth to it, but it means much less than is commonly understood.
It's refreshing to read human prose about something like this. There's so much you can get about the author's state of mind -- pride at what they've accomplished and how they like to conceptualize it, traces of how they learned things and what those things mean to them... for me, it's a reminder of what we've lost in giving over so much of our writing to AI.
But it's understandable that the writeup would trigger defensiveness. It claims a lot of originality and novelty that seem overreaching. AIUI, it's an optimizing JIT with a superoptimizer and a shared cache, with the search procedure guided by AI. And the equivalence model that the superoptimizer is based on includes all of Python and can handle both CPU and GPU code. It's quite impressive, really.
It's hard to tell what of it is new. Superoptimizers are not, JITs are not, caching JIT code is not, using machine learning to speculate or search is not. Constraining an LLM to produce correct output is no newer than forcing semantically valid JSON output. (You could think them as different type systems.) Different JITs have handled different kinds and amounts of runtime information -- tracing JITs, runtime type inference like SpiderMonkey used to use, speculative compilation and invalidation (global or just in time like inline caches)....
But none of that takes away from Maxwell. There isn't really enough information to compare it to other systems. Random example: many JITs work really hard to balance compilation time with execution time, and the post doesn't talk about that. Is it only for long-running code? Does it reach a steady state, and if so how much overhead is there at that time? What does it make slower or bigger?
At any rate, it seems like there's a lot of interesting work in Maxwell, and it's worth checking out.
It's something, at least.
I now try to only give examples of how I want it to do things, and hooks that complain when they do the bad thing (but still don't mention the bad thing in the complaint.)
Email was awesome when it first became widespread. I could cold-email random professors at other universities (I was in school at the time) and get meaningful responses back. Then lots of people (and companies) started cold-emailing, and it stopped working.
Social media is broken if there's nobody there, is great with a medium-sized community, and is broken if everyone is there.
Ad-supported websites can offer all kinds of valuable free services. Up until either (1) they become so ad-infested that they're no longer useful, or (2) all the users block ads.
There was a wave of websites providing APIs for programmatic access a while back. You could script together whatever flow you wanted. If no API was offered, you could scrape the HTML. That triggered a wave of alternate interfaces to sites, skipping their ads, skipping their engagement funnels, separating the value from the cost. So everything got locked up behind captchas and accounts and API keys with limited access.
This outcome was always inevitable. Humans don't understand social contracts, or they do and both sides abuse the hell out of it. We feel entitled to have just the loss leaders without ever buying anything else. When running businesses, we feel entitled to keep adding more and more adds and generally enshittifying the experience. People can take it or leave it, after all.
Agents don't change the underlying structure at all. They just magnify and accelerate parts of the system, and will therefore magnify and accelerate the reactive and defensive responses.
"Anybody who wants to" is precisely the problem -- there's no correlation between a rando's wishes and the health of a project or working environment. You need people with skin in the game, and in the right game.
But they should be marked experimental in that case, not optional. Optional flags ought to work together, or at least it should be considered a bug worth quickly fixing if they don't. Combining experimental flags otoh is Buyer Beware.
