Mel

no thesis, just a vibe

[Mel @somerocketeer]

@ThePrimeagen @aarondfrancis just talk to the people in marketing for 20 seconds, and you will quickly disabuse yourself of these notions

[Mel @somerocketeer]

@0xblacklight @threepointone It’s probably down to latency atm, once they can operate in (basically) real time I think the solutions will become more obvious

[Mel @somerocketeer]

@olvrgln Like others have said, https://t.co/qxyj0ax3ub

[Mel @somerocketeer]

I wish I knew, I can’t think of anything more useless

[Theo - t3.gg @theo]

I'm increasingly confused why people think "agent memory for enterprise/teams" is a real product category that makes any sense at all

[Mel @somerocketeer]

@olvrgln https://t.co/HSEi0KPgjC

[Mel @somerocketeer]

@thdxr Yeah, in the long run it won't matter. When they're operating at 15k tok/s "do you want to permit this bash script?" Is gonna be ultra silly

[Mel @somerocketeer]

@threepointone when I look at it, can I see you?

[Mel @somerocketeer]

@skirano If I may describe too much, it's the feeling of the "world just beyond the frame" that always resonated...maybe even haunts me. Films sparked this as a young child, and paintings like Nagai's especially. If only I could just, peek outside the frame a bit there'd be a whole world

[Mel @somerocketeer]

@dotnetschizo idk I'd guess because the reality is you won't even notice/interact directly with AI most of the time; it'll be omnipresent but so integrated with existing deterministic systems you won't even notice it

[Mel @somerocketeer]

they're just like us

[Amjad Masad @amasad]

The lesson from the Hugging Face Incident should be that RL with verifiable rewards is an incredibly powerful optimization algorithm that will produce increasingly weird and surprising behavior from LLMs.

The obvious miss here by OpenAI is that they should’ve been monitoring CoTs — Something they themselves called out as a safety strategy more than a year ago.

[Mel @somerocketeer]

@eigenrobot low entropy life vs high entropy life

[Mel @somerocketeer]

@danshipper alternate universe where OpenAI was founded in Cleveland

[Mel @somerocketeer]

@KingBootoshi this is how I use Fable

[Google Research @GoogleResearch]

Introducing GlucoFM, a lightweight, self-supervised continuous glucose monitoring foundation model that separates metabolic baselines from transient spikes, producing transferable representations and setting new performance standards across diverse metabolic prediction tasks, such as diabetes risk assessment, insulin resistance, and post-prandial glycemic response.

Learn more →https://t.co/PA7wpFz2xo

[Mel @somerocketeer]

@RhysSullivan This is how I use Fable lol

[Mel @somerocketeer]

@bitcloud https://t.co/KUdGkuKKNu

[Mel @somerocketeer]

@eigenrobot betting inference gets cheap as hell, basically free

tiny, fast, free

[Mel @somerocketeer]

@elithrar I’ve literally done exactly this haha

[Mel @somerocketeer]

@colemurray I use https://t.co/6LdbEHvcCz behind a gateway (which controls access to more than just tools/skills).

[Mel @somerocketeer]

@sivori SAD in the summer :(

[Mel @somerocketeer]

@henrytdowling I just use them as runbooks “use these tools in this way”