maherbeg

I think all of these companies need to let you separate secrets from other env vars and not let you ever read your secrets once they're written down.

These companies probably have so much data on stuff that failed that there are likely so many gems and paths we haven’t tried out yet

I don't think we're in a bubble yet for AI. We've only really reached the software engineering market, and even then it's not permeated throughout every software engineering firm. There are so many other engineering disciplines to accelerate!

I wonder if the US will actually be fine here. We're starting to restrict immigration (at all levels, legal, illegal, skilled, unskilled) and boomers are retiring en-masse. Will the productivity gains negate the loss of labor. Labor isn't quite fungible but we might find some equilibrium between these 2 functions.

I think we'll end up seeing legislation against the license version of all of this content. Instead, you'll get to own it and move it around as necessary. This hasn't been necessary yet, but if that becomes the only option, then ownership becomes a more important question.

Which nuc were you running on? I've got an 8th gen that I'm also considering upgrading but wish I could up to way more RAM.

Pretty sweet. I wonder if they can continue to distill this down so it runs smaller. I think we're at a point where more intelligence isn't really needed for a lot of use cases.

Also honestly, the models don't have to get much smarter to handle the consumer side of things. They could innovate a ton on the harness to farm out requests and manage the state of the agent. They have an incentive to do this to reduce the infra usage so they can go wider with the agent.

I agree with this. They should have to keep some insane bond for every phone number they keep to pay out when there is a data leak.

This is so awesome. I'm hoping Apple does the same.

This is awesome! Please submit updates when a new model comes out that people are saying are better than Jev!

Has anyone actually eval'd the other open source options against Jev on real world tasks rather than looking at benchmarks?

I see a lot of people parroting the quick open source alternatives as being better on the benchmarks, but it's such a new category that I'm not convinced we have solid benchmarks.

I'm hoping a company releases an internal eval benchmark for these options. I'm sure some of the open source ones are solid in some cases, but would love to see more reliable data.

Slack? I think some of their plugins API was all Deno based for a while.

Love it. I've been using $visualize a lot in the codex desktop app, and having even richer experiences will be sweet.

We just need an LLM based Lisp so we can finally close the loop of code is data and data is code.