[basisword]

It would definitely be a problem. A big one. Think of all the multi-national tech companies that have rolled out AI not only to their engineering teams but to their business teams now too. Suddenly your employees in the US can do more and do it quicker than your employees elsewhere. It would be a nightmare to try and manage.

[graemep]

That only applies to things Mythos does better than Opus. Mythos is supposedly very good at things such as finding vulnerabilities, but is it better at business tasks? IN the short time I had access to Fable it did not seem noticeably better at things I tried it for (lots of small tasks).

Maybe it is better at vibe coding or finding security flaws, but at how much is it sufficiently better to be worth paying the extra?

[basisword]

You're assuming this export restriction won't apply to any future models. The problem is a couple of years from now when the ROW is still on Opus and US workers are on something much better.

[graemep]

Which would force us to use Chinese models instead. It would also increase the incentive for other countries to develop models.

[grumple]

I disagree. I work at a very large international corporation. How much revenue do you think we’ve seen due to AI? I’d guess it’s zero. I know for the groups whose finances I see, it’s zero. Yet costs have gone way up. There’s a bunch of new code, but not anything customers are going to pay more for.

And either way no AI basically puts you back to where we were last year. US employees have always been far more productive, that’s nothing new.

[basisword]

>> There’s a bunch of new code, but not anything customers are going to pay more for.

This isn't an AI problem, it's a management issue. Why are you getting devs to build things that aren't going to generate revenue?

[grumple]

Do you think the only bottleneck in dev has been the speed of coding? I think it’s obvious that this is not the case. It’s finding product market fit, actually discovering and deciding what must be built, and then selling it. And there’s a deeper economic reality: budgets for software are finite and limited, and already close to maxed for most consumers and businesses. If my customers can only afford $500/month total for software, no amount of software I make will push them past that.

It’s not just us: where’s the revenue in the entire market? We can see all the public filings. There haven’t been any revenue gains. The only people making money from AI are the LLM providers. And even they are losing money. Even the biggest tech companies are limiting token spend. At best the tech is a new cost just to maintain parity, and I think most businesses look at it as a way to cut dev costs (trading for token spend). I think they will learn that it’s less of a win than they hoped. If a dev was spending 50% of their time coding, and you reduce that to 10% - that’s a big change but it isn’t really making you more because it’s all that time we hate in meetings, understanding customer needs, etc, that make us money.

[Schiendelman]

A lot of the bottleneck is product management now. There's always near infinite work in every engineering team's backlog.

Identifying the next few things to improve the product is hard work, with market investigation, competitive analysis, synthesizing user feedback... some of this gets faster with LLMs, certainly.

I think most tech companies haven't figured out how to rebalance product to engineering to ensure engineering teams are doing work that grows the business funnel rather than expending tokens on other work.