Hacker News: jasonjmcghee

New comment by jasonjmcghee in "Claude Duet"

jasonjmcghee — Wed, 21 Oct 2026 19:30:23 +0000

I feel like the link to be used here should be https://claude.com/resources/articles/claude-duet instead.

New comment by jasonjmcghee in "Any window from your Mac, on your iPhone and iPad"

jasonjmcghee — Fri, 16 Oct 2026 02:25:35 +0000

My guess is this is meant to be a no friction VNC (though it doesn't seem to be using VNC). So you can be away from home and remote in.

There are VNC clients on iphone / ipad - but generally it's whole desktop.

RealVNC is the name brand but it's not exactly a good experience. "Screens 5" looks like a better alternative.

New comment by jasonjmcghee in "AlphaNIST"

jasonjmcghee — Wed, 14 Oct 2026 01:13:14 +0000

Fair - I should have

New comment by jasonjmcghee in "AlphaNIST"

jasonjmcghee — Wed, 14 Oct 2026 00:14:54 +0000

NIST makes sense, but co-opting Alpha is a bit odd.

Alpha historically had meant "Google DeepMind" not RL

New comment by jasonjmcghee in "AlphaNIST"

jasonjmcghee — Wed, 14 Oct 2026 00:07:58 +0000

It's the companion blog post to an Nvidia / UW paper of an approach to solving learning based high precision robotics tasks.

HN guidelines are to use the title of the blog post and not editorialize.

AlphaNIST

jasonjmcghee — Tue, 13 Oct 2026 21:57:29 +0000

Article URL: https://alpha-nist.github.io/

Comments URL: https://news.ycombinator.com/item?id=50079464

Points: 98

# Comments: 38

New comment by jasonjmcghee in "AWS Lambda Web Functions"

jasonjmcghee — Tue, 13 Oct 2026 13:00:39 +0000

And there are even hibernateable websockets too - only paying for when data is actually transferring.

And the maximum worker time is based on processing time, not wall time.

New comment by jasonjmcghee in "Show HN: BetterWispr – Free, open-source dictation for Mac"

jasonjmcghee — Sun, 11 Oct 2026 15:55:03 +0000

The native / built-in API SpeechAnalyzer (introduced last generation on iOS and MacOS) works quite well.

Are folks finding it lacking / needing alternatives?

New comment by jasonjmcghee in "ESP32-C3 Adblock"

jasonjmcghee — Wed, 07 Oct 2026 07:15:52 +0000

I sure hope that's not true. Maybe hit the comments section first?

New comment by jasonjmcghee in "Jev-Driven SRE Diagnosis: What Worked and What Failed"

jasonjmcghee — Wed, 07 Oct 2026 06:32:20 +0000

I'm guessing you're getting downvoted due to dismissive tone, but I think there's truth to this.

Labeling something as "don't escalate" that should have been isn't great. Paying 5x and having to wait 10s instead of 100ms or whatever for a reasoning model (with tools?) is very likely worth it.

New comment by jasonjmcghee in "Decisions API is in public beta"

jasonjmcghee — Wed, 07 Oct 2026 00:30:21 +0000

Different APIs for different things reminds me of the early auto-complete vs instruction apis.

Will this get folded into models / post training pipelines at some point and make them better at calibrated outputs?

New comment by jasonjmcghee in "Decisions API is in public beta"

jasonjmcghee — Wed, 07 Oct 2026 00:28:23 +0000

Or even just- they've already procured OpenAI.

That's a big motivator.

New comment by jasonjmcghee in "Mistral Large 4"

jasonjmcghee — Tue, 06 Oct 2026 14:05:01 +0000

Had the same thought - feels chart crime adjacent

New comment by jasonjmcghee in "Web Search API"

jasonjmcghee — Mon, 05 Oct 2026 14:02:40 +0000

> All three support Zero Data Retention for requests made through Cloudflare

And then on the providers page:

    Property              Value
    provider              exa
    Zero Data Retention   No

New comment by jasonjmcghee in "From the creator of Redis; run LLM locally with ds4"

jasonjmcghee — Sat, 03 Oct 2026 03:14:38 +0000

Last time I was using llama cpp you could just do:

    llama_state_save_file

    llama_state_seq_save_file

and the load equivalents.

That was a year or so ago though...

New comment by jasonjmcghee in "Muse Gadgets"

jasonjmcghee — Sat, 03 Oct 2026 00:01:30 +0000

This and the Meta VR stuff - I would be all over it if it weren't Meta. It's objectively cool tech.

But you're handcuffed.

New comment by jasonjmcghee in "Micron CEO Says Memory Supply Will Be Much Tighter in 2027 and 2028 Than in 2026"

jasonjmcghee — Thu, 01 Oct 2026 13:57:21 +0000

Exactly - this kind of borders on market influence.

Effectively: "Now's the time to invest..."

New comment by jasonjmcghee in "Dear Software Makers"

jasonjmcghee — Thu, 01 Oct 2026 13:55:30 +0000

> YouTube spends way too much time chasing their competitors

Is A/B testing a thing on TikTok and/or Instagram (or similar platforms)?

New comment by jasonjmcghee in "Gemini 4 Argon"

jasonjmcghee — Wed, 30 Sep 2026 20:41:00 +0000

Gemini 1.5 Pro claimed 10M input tokens before release.

And was 2M tokens IIRC after release.

There were also many rumors that Gemini 4 was going back to 2M. Just seems odd not to say what it is.

New comment by jasonjmcghee in "Gemini 4 Argon"

jasonjmcghee — Wed, 30 Sep 2026 20:12:21 +0000

> 1M output token limit

what about input?

(Maybe I missed it)