That's really nice. Have you tested if it works well with longer and more detailed prompts? For example adding more whole product specs and so on. It would be nice to generate a design system from generated UI you like instead of recreating that UI directly.

RE longer prompts, yes. Generally speaking I expand most prompts to be around a full page of text as it is already, so adding more detail in just refines that expanded prompt. That's more for a single screen though. It sounds like what you're asking about is something like a design.md for an entire brand / docs for a design system.

For that, I have a different approach, which is to extract your design system from screenshots. After which you can just select the brand you want when generating. There's sample images in the sibling comment in this thread.

Also it might be worth noting my pedigree here - I ran the design systems features over at Figma for around 5 years, but quit to build out diffui. The project is heavily oriented towards being able to replicate brands consistently, since the target audience I'm going after is enterprise design teams who are having trouble with existing tools capturing their brand look/feel.

If the frontier models will take as much money to train as they do now, there is no way the wealthy are able to afford their training just for their own consumption. Financing of this whole thing rests on the models being available to companies and consumers who are willing to pay astronomical (compared to other software) sums for it.

This is a good point. If I were an investor, there is no way I'm investing into frontier labs after this announcement. Is this how the bubble pops?

No, there is still money to be made in self hosted LLM models. You invest in the AI infrastructure companies, not frontier labs. The bubble continues.

Moreover, is this a clever way for the frontier labs to try and “dodge” the negative impacts of the bubble popping.

“Oh no it’s not our fault!!1! ~~My dad works at Nintendo~~ the govt told us we weren’t allowed to release it”

It could also be something PSD2 based. You should be able to create payment using PSD2, but the client still has to approve it inside their bank app.

They said Australian - it's POLi.

It was indeed POLi. Strange that it's using actual bank passwords!

OpenAI & Anthropic are winning right now. I suspect if Chinese companies get ahead in the race the cards will reverse, OpenAI will restart farming goodwill with open models and then winning companies will be releasing closed models.

Besides points already mentioned,

- remote mcps are server driven, meaning the producer can introduce new functionality without requiring all clients to update their skills and clis

- remote mcps are safe as they don't require literal code execution privileges on your system. Many times skills even bundle scripts with `npx`/`uvx` which is basically just `curl npm.com | bash` level of unsafe

Wow that looks really impressive. Both the UI and the content looks good, the game is a bit buggy but still nice!

With AI dependence, unless you are a holdout, offline development isn't really a thing anymore. Perhaps to do some code reviews, but actually producing new code?

That is sad.

Somewhat, it's a tradeoff we are making willingly

I haven't made that tradeoff. I still code myself without agents from time to time.

Nothing changed. You can still code the old way. All those 100x productivity gains are probably closer to 10% productivity gains after you account for all the added debugging and steering.

Doordash and similar are experimenting with autonomous/remotely operated vehicles and porn is getting decimated once good enough uncensored video gen ai gets available. That doesn't sound like viable career choices either.

You can totally get a random "we've detected a suspicious activity, please verify the phone number you've used during registration" form during login.