Hi ๐ I'm Paul. Eons ago I left a flourishing quant dev career to build the games I loved playing as a kid.
First stop: Hexfell โ a Dorfromantik-inspired deckbuilder about growing tiny island landscapes, with a deck of Tiles.
Here's a sneak preview ๐ #indiedev#gamedev
Wondering how soon we'll start seeing the effects of the interval vs external model divergence. Independently of the reason for doing it, not shipping the same product internally and externally can quickly make it more difficult to build the right thing.
This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards. The benchmarks are great and it's SOTA on everything by a margin but I'll add that *qualitatively* also, this is a major-version-bump-deserving step change forward (imo of the same order as Claude 4.5 was in November), peaking especially for long problem-solving sessions on very difficult problems. You can give it a lot more ambitious tasks than what you're used to, the model "gets it" and it will just go, and it's never felt this tempting to stop looking at the code at all (but don't do this in prod!). The model still has quirks that people will run into and the safeguards are configured to be a little too trigger happy for launch, which can hopefully be tuned over time.
I feel a lot of things changing as working software increasingly comes out on a tap. The Jevon's paradox kicks in and I feel my own demand for software growing substantially. You can ask for anything - explainers, visualizers, dashboards, bespoke single-use apps (e.g. a full wandb that is hyper-specific just for your project), you can 10X your test suite, auto-optimize code, run giant research projects with custom HTML for the results, anything! "Free your mind" (Matrix ref). Really looking forward to all the things people build!
@lillyemira@UseCorgi I think the point is more around the performative external discourse - many founders do 997 but the expectation you should be setting with your team is the output: I think it's more reasonable to set crazy ambitious goals than to say you expect people to sleep on the office floor
I'm constantly amazed by how well @SpokenlyApp works with the local Parakeet model - these days a fairly old M1 runs it effortlessly and it's just so much better than using dictation in the frontier harnesses.
@cosminnegruseri Companies are bringing back in-person interviews, so I suspect leetcode will get a second life. Whether it's a good signal for hiring is a different question