Sure, but I think we should judiciously avoid the false equivalence yielded by only looking at this on a developer-by-developer basis, rather than systemically. The truth is that in practice, AI is not a neutral force. Obviously AI can enhance the output of smart, experienced developers and improve the efficiency of code reviews, mitigating the effects of garbage PRs. However, it increases the percentage of PRs contributed by entirely inexperienced and/or not-smart devs from zero to, potentially, the majority. It entirely removes the barriers inherent to coding that kept Dunning-Krueger cases from submitting ill-conceived or poorly constructed changes— actually getting them to run in some way, even poorly. That makes them much more difficult to distinguish from well-constructed PRs than those from, say, someone cargo-culting code from tutorials.
Moreover, as these tools become more expensive, people with money to blow on tokens will be able to drown maintainers that don’t have enough token-cash to help them deal with it. People see this as mostly a matter of time and energy, but I reckon it will soon be a financial issue.
I see AI as a barrier remover. Unfortunately some barriers are good or minimally necessary.
I think we'll need to revert to artificial barriers such as bonds, e.g., if you want to do a PR to my repository you need to pay a 10 dollar bond. If the PR is good and I want future PRs, you keep your bond. If it's slop and spam, I get 10 dollars for my time.
This is entirely too much friction in the wrong place. Public open source will simply die before a system like that ever becomes the norm.
The previous barriers worked because they were organically perfectly in line with a contributor's internal incentives. A contributor gains very little benefit from submitting a patch; the likelihood is infinitesimally small they'll ever get any career advancement, financial recompense, or even much community recognition for it. At most, it shifts the burden of maintaining the code they're contributing from themselves to the community / long-term maintainers. The real incentive for a contributor was making the patch, because they get to see the feature or fix they want made for the software. The previous barriers were in making the patch, and contributors would overcome that friction to gain the benefit of having the patch they want. Moving the barrier to merely submitting the patch after it has already been made will simply result in people not bothering, because there is very little incentivizing them to deal with the friction.
1) more reliance on systems to track reputation across projects. I'm sure Microsoft, in the form of GitHub, will love to sell you a partial fix to the same problems it so enthusiastically helped to create. But there are the familiar problems of surveillance, identity theft, office politics, and system-gaming, and it doesn't on its own offer an onramp for new players.
2) in-person coding tests at the same Pearson test centres where people take most of their Cisco (and accounting, and ...) exams today. Not as expensive or inconvenient as you might think, but not the cheapest and easiest, and it certainly has the same concerns re. surveillance and identity theft
I agree with the bond in theory, but that would entirely stop contributions from people in economies where a shady maintainer could keep their code, and their weekly food budget.
We already have trouble with people maintaining open source projects without getting paid, now you want people to pay for the privilege to participate in free work?
Same ratio of imbalance, just with matching multipliers distributed to each side, and everybody is probably worse off because of it: I cite post-LLM-ATS hiring/job hunting.
> They can only do that if they're a monopoly, which they're not
Why do you say that? I reckon lots and lots of companies sell software that aren’t monopolies. Having competition, even stiff competition, isn’t anathema to running a business.
Just to clarify, I’m not the person you initially replied to.
> "They wouldn't be selling tokens directly ... They'd hoard them" But they can't do that because they aren't monopolies.
Hoarding them— not selling any of them, but instead using them internally and selling the products created by them — doesn’t at all seem like it would require a monopoly.
I think most people that dismiss BEAM right off the bat either don’t understand the built-in beam process/supervisor/etc. model with its inherent fault tolerance, etc, or assume its not useful because it doesn’t address their use cases.
That or it’s a evangelist from the church of AI speaking based on faith rather than reason.
Much less emasculating than accepting that all that weird tech mumbo jumbo that those overpaid senior engineers babble about actually has a deeper meaning and is not just there to artificially slow down the all important company growth!
Google can’t spend two decades getting bazillions of people to rely on what’s essentially internet infrastructure at this point, and then pretend their hands aren’t dirty when they suddenly crank the enshittification juicer up to 11 because they need to justify the gobsmscking capex for their largely hated new service. There’s nothing legally stopping them from doing that, but that’s very different than right and wrong.
I am actually up all the drinks I got comped in Vegas. I sit down at the penny slots and bet one penny one row until I get offered a free drink. I tip the server $3, bet two more pennies for good measure, get up, and walk out with the drink in my hand. I just got like a $3.10 Manhattan for walking around the strip, including tip, courtesy of some business that was low-key trying to scam me and deserves to have less money than they do.
I worked in the bar business and know what a shot of rail bourbon costs with a volume discount, and the ten cents they get from the slot machine definitely does not cover it. I’m happy to give the server $3 for their effort.
They’re obviously doing it because it’s profitable. If they were being genuinely altruistic, I’d be much less likely to take advantage of it.
So someone could go and teach a class on how to build pipe bombs, refine ricin, shake-and-bake meth, 3D print guns, and all sorts of other things like that, and when the ATF looked into it, they’d just be like “well technically this is all out there on the Internet, in library books, etc. Guess it’s ok!”
This suit has nothing to do with free speech and the F1A provides no relevant protection here. This prosecution is under consumer protection law. Broadly, the cause of action is "you negligently sold a defective product which you knew (or should have known) actually causes harm or is likely to cause harm." Proving negligence (willful or otherwise) depends significantly on things like the sales and usage context as well as claimed features of the product along with disclaimers, disclosures, existing practice, prior knowledge of actual harm, etc.
That the product or service in question included supplying information that was publicly available elsewhere wouldn't be an effective defense against claims of willful negligence or reckless endangerment. For example, rat poison is sold in in certain retailers in packaging with copious warnings and successful prosecutions under product liability or consumer protection law are rare. But if another company sold rat poison in bright pink boxes with a cute cartoon mascot and no warnings in toy stores - and then kept selling it after they knew three children had bought it and died - the fact the same chemical compound is also commonly sold in hardware stores wouldn't be relevant.
To win a judgement, the AG will need to prove that ChatGPT was clearly a dangerous product and OAI acted negligently in supplying it to customers it knew (or should have known) were vulnerable. This will be quite a stretch under existing law. I suspect the AG has no intention of taking this case to trial and, shortly after the November elections, will settle for a lump sum fine paid to the state treasury and a vaguely worded consent decree which mirrors internal policies and product changes OAI has already adopted to minimize liability.
Moreover, as these tools become more expensive, people with money to blow on tokens will be able to drown maintainers that don’t have enough token-cash to help them deal with it. People see this as mostly a matter of time and energy, but I reckon it will soon be a financial issue.
reply