In my experience personal projects are the greatest indicator of IC competence, especially for young people. You may not like it, but turns out that when you do a thing in your free time because you like it, you get better at the thing than the people that only do it because they have to.
Personally I really dislike when the agents generate super long composed shell commands because they are really hard to audit. ffmpeg I'd whitelist, but if it makes a mistake in some super long chained git command it can have pretty scary consequences.
That's really nice. It would be really good for game GUIs too where the situation is quite poor and would work well with underlays/overlays/worldspace UIs. That said while binary size may be around 10mb, it still baloons to 500mb at runtime for your TODO list example which is more than some electron apps.
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.
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.
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.
- 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
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?
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.
I have been thinking about this a lot lately. If you look at effects of geopolitical events and who profits and loses, rather than stated intentions. This global oil crisis, the Ukraine crisis, tariffs ect. It's the equivalent of the "the purpose of the system is what it does".
Oil crisis: Trumps friends profit on insider info, US oil industry (also his friends) profits, Russia profits because they are another big oil producer, USD dominance is harmed (also helps Russia), everyone else in the world eats the costs
Ukraine: Russia bleeds, Ukraine bleeds, arms industry profits, politicians in general get something to grandstand on in front of the voters. Personally I believe that this conflict has been artificially prolonged just to amplify the effects
Tariffs: US public eats the costs, Trump profits politically by appearing strong, Trumps friends profit on insider info
I don't really get how this stops captcha solving as a service, which is the actual way that scaled recaptcha solving is done? Those things are incredibly cheap and are staffed by humans anyway. Instead of selecting grainy busses, they will just scan the image with their phones.
Perhaps for some very specific capabilities such as TTS, translation, voice recognition and so on. But for general intelligence models, better hardware just directly allows better models and that doesn't seem to be changing any time soon.
My point is that since we have had so few nuclear incidents, but they have done massive damage, it is very possible that we don't actually know much worse it could get. We have only seen a few points from a distribution that could be much wider than we think. Compared to renewable failures for which we have a pretty good idea.