aliljet

There's a vibe in at least the PNW that feels like the tech sector is sloughing jobs and avoiding creating new ones courtesy of AI. I genuinely wonder if that feeling is backed by reality and whether it's large enough to be translating into national statistics across all industries.

Does fine tuning really improve anything above just pure RAG approaches for usee cases that involve tons of direct document context?

This is really what I want to understand. Where can we see real world performance benchmarks?

Is this actually true? I want to see actual evals that match this up with Sonnet 4.5.

How do you even pull away from a Gmail address? I'm nearly twenty years into that service. Getting banned would be absolutely devastating...

I really really want to see how these images are starting to form into videos. The stills are clearly getting better and better, but what about when you need the stills to organically conform to a keyed script?

The easiest way? Angelist.

The path to declaring yourself accredited is uniquely easy. Just say it. The whole space is deeply unregulated and unaudited. What makes it insane is that those middleman are making a small fortune exploiting this loophole protecting large companies from being forced to go public. The number is 2000 private investors. Rest assured, more than 2000 individuals have money in Stripe today. It's a total scam.

The public can absolutely participate in this by way of syndication deals. Those syndicates are what's covering up the true extent of ownership and they're essentially charging for access with their fees. It's oddly shady, poorly regulated, and more expensive than just being public, but everyone can ride this ride.

That's a bit confusing. Do you believe LLMs coming out of non-chinese labs are censoring information about Israel and/or Palestine? Can you provide examples?

I wonder if this will extend SEV-SNP and TDX to the child VMs?

I wonder if these are starting to get reasonable enough to use locally?

The problem here is that it looks like this is released with almost no real access. How are people using this without submitting to a $250/mo subscription?

What I really want is for the peon voice to be replicated and for custom things to be in that voice. Or even better, the starcraft battlecruiser guy's voice!

Am I just a cynic, or do any of the LLMs deserve love too?

This is precisely the real question. If you're exceeding human transcription, you may be generally pretty good. The question is what happens when you tell a human to become surgical about some part of the document, how then does the comparison change..

All of healthcare is crying. Trust me.

This is actually the thing I really desperately need. I'm routinely analyzing contracts that were faxed to me, scanned with monstrously poor resolution, wet signed, all kinds of shit. The big LLM providers choke on this raw input and I burn up the entire context window for 30 pages of text. Understandable evals of the quality of these OCR systems (which are moving wicked fast) would be helpful...

And here's the kicker. I can't afford mistakes. Missing a single character or misinterpreting it could be catastrophic. 4 units vacant? 10 days to respond? Signature missing? Incredibly critical things. I can't find an eval that gives me confidence around this.

I wonder if we're going to see an aurora over Seattle tonight?

Is there clarity right now around foreign students attempting to obtain h1bs in the future?

This is absolutely fantastic. I wish this included commercial loans like DSCRs...

I wonder how effective this is medical scenarios? Segmenting organs and tumors in cat scans or MRIs?

The real question I have after seeing the usage rug being pulled is what this costs and how usable this ACTUALLY is with a Claude Max 20x subscription. In practice, Opus is basically unusable by anyone paying enterprise-prices. And the modification of "usage" quotas has made the platform fundamentally unstable, and honestly, it left me personally feeling like I was cheated by Anthropic...

This is great, but I was hoping to read a bunch of hilarious poetry. Where is the actual poetry?!

Thus far, this is one of the best objective evaluations of real world software engineering...

Understanding precisely why Gemini 3 isn't front of the pack on SWE Bench is really what I was hoping to understand here. Especially for a blog post targeted at software developers...

This is the heroic move everyone is waiting for. Do you know how this will be priced?

When will this be available in the cli?

What's wild here is that among every single score they've absolutely killed, somehow, Anthropic and Claude Sonnet 4.5 have won a single victory in the fight: SWE Bench Verified and only by a singular point.

I already enjoy Gemini 2.5 pro for planning and if Gemini 3 is priced similarly, I'll be incredibly happy to ditch the painfully pricey Claude max subscription. To be fair, I've already got an extremely sour taste in my mouth from the last Anthropic bait and switch on pricing and usage, so happy to see Google take the crown here.

What we really desperately need is more context pruning from these LLMs. The ability to pull irrelevant parts of the context window as a task is brought into focus.