neilellis

Claude had munchies....

Did you actually try them, I tried a sample they were rubbish compared to Jev - I'm not saying they will survive, but the competition CURRENTLY sucks - and also they aren't actually cheaper if you need them at scale. OpenAI is coming in twice the cost and so is Cloudflare. Ad yes you can finetune, and yes I do, but it also SUCKS. Fine-tuning and managing your own datasets is yet another timesink. So we'll see, but it's not quite black and white.

So they ban us from scraping sites but they can do it themselves, this stinks. How did we let them become the internets gate keeper.

The Onion isn't satire it's just a dystopian Cassandara.

Well now all the agents will have this in their training data, so expect them to come calling in the future :-) :-)

Peak HN comment :-)

Easy for you to say.

I like it, very smart for agents at scale.

'frontier models' - seriously, it was you and only you!

Good luck with this model/product, in the excitement of LLMs people seem to forget applicability. I very much like to see innovation in this space, so well done!

Also FYI doesn't run inference on Apple GPU (only for training)

I tried this today for labelling - and for that task it was very bad MNLI was better - so you are going to need to match the use case for this pretty exactly. (at 29MB params one would expect that!) I'm obviously not saying labelling is a good use case :-) just adding a data point.

Jev has put the cat amongst the pigeons so suddenly everyone is looking at classifiers and encoder only models again.

My ideal model would be a general purpose LLM API that can answer classification questions and as it does so distils to an encoder only model so that the more classifications I do the cheaper it gets (i.e. the more it offloads to the classifier). If anyone ever wants to do this as a service do let me know, because it's just another piece of code to manage in each new project that needs classification.

Also a model that could do this internally would be nice :-)

Please have a 'readable version' option so I don't have to exhaust myself parsing the sites layout. I get that it's unique but most of us just want to work out what you're offering in 5-10 seconds of our time.

OMG blast from the past!

Correct me if I'm wrong but Jev itself works pretty much the same as encoder only models.

Well that benchmark is now saturated, what next. How fast you can hack the pentagon?

There are no shortage of good harnesses out there, and you can re-use coding ones like pi, opencode, dsh even claude and codex. For most use-cases that is already the loop you need. If you're looking for something less like a coding agent then Vercel's Eve is okay too.

Why on earth would we want such a lock-in at this stage when there is no clear winner. This is an area in which I would encourage everyone to build their own (using OSS) on top of existing cloud infrastructure.

Congratulations!!! You win what’s left of the internet - just ask Claude for your prize! Motrin I’ve had this week.

I use codex now.

Aliens playing D&D

I'll get around to reading this at some point.

Dystopic

You guys still having elections?

Also relevant: Ten Percent Of U.S. High School Students Graduating Without Basic Object Permanence Skills - https://www.youtube.com/watch?v=ssjokgx0pUQ

Unless it’s already happened and we missed it

Caught in a landslide, no escape from reality

When I hear people talking about how insecure OpenClaw is, I remember how insecure the internet was in the early days. Sometimes it's about doing the right thing badly and fix the bad things after.

Big Tech can't release software this dangerous and then figure out how to make it secure. For them it would be an absolute disaster and could ruin them.

What OpenClaw did was show us the future, give us a taste of what it would be like and had the balls to do it badly.

Technology is often pushed forwards by ostensively bad ideas (like telnet) that carve a path through the jungle and let other people create roads after.

I don't get the hate towards OpenClaw, if it was a consumer product I would, but for hackers to play around to see what is possible it's an amazing (and ridiculously simple) idea. Much like http was.

If you connected to your bank account via telnet in the 1980s or plain http in the 90s or stored your secrets in 'crypt' well, you deserved what you got ;-) But that's how many great things get started, badly, we see the flaws fix them and we get the safe version.

And that I guess is what he'll get to do now.

* OpenClaw is a straw man for AGI *

Yes, also for semantic indexes, I use one for person/role/org matches. So that CEO == chief executive ~= managing director good when you have grey data and multiple look up data sources that use different terms.

That I would like to see too, usearch is amazingly fast, 44m embeddings in < 100ms

It's ahead in raw power but not in function. Like it's got the worlds fast engine but one gear! Trouble is some benchmarks only measure horse power.