Two fun things about summarize in click to do: - it moderates, very possibly using the cloud, so text may not actually stay on device. Microsoft moment - It has a very weak mind and can trivially be convinced to do other things. May or may not be phi-4-silica

You're absolutely right!

IT'S HERE!!!

Swear to fucking God Claude code web just had "yeah nah mate" as it's status text

Dell latitude 7200 detachable I got for like 150$ has 36% battery health 💀 ordered a new one for like 30 dollars. Case is also cracked but I can fix that later.

Hold on, I was thinking, why didn't they use one of snapdragon's X series chips here? Could've had like, Intel ultra 7/9 performance for not a super whole lot more power draw? Seems like a missed opportunity!

new episodes of dispatch are out today

The frame runs Linux!!!!! The steam frame runs Linux!!!!! Standalone Linux VR headset!!!

It's real!!!

I have reverse engineered text-dollcode! Or rather, Claude did, in a single attempt. Written in C#, also supports custom symbols! github.com/Isolyth/Doll...

One of my favorite LLM use cases is literally just asking it what the closest preexisting concept to a thought is. "I have this idea, has anyone had this idea before and what's it called?" It used to be REALLY DIFFICULT to reliably answer questions like that and I feel like we'll forget it was.

Anyone know what happened to github.com/v01dlabs/dol... and @demi ? I don't have an archive of the code so I'm not sure how I can decode (or encode again!) the dollcode in my bio :(

When I started my doctorate, I had to complete the standard research ethics training. Particularly with prisoner populations, I saw many parallels to AI. #ai #artificialintelligence #llms #largelanguagemodels #aiethics

no, they're weirdly coy about this they produced a 235B, 480B, and then 1T model within a span of like 2 months though and there are some artifacts in the 30B this is also the subject of enduring rumors which I have seen reasonably trustworthy people vouch for

PSA: Some people got emailed about this and some did not, but everyone with a Claude pro or max plan got $250 or $1000 worth of free credits to use on the web version of Claude code, it's valid until the 18th

Claude has memory on the pro tier now.

After lots of images, an example project, and giving it lots and lots of praise (I think this is what did it) Claude was successfully able to add concave corners to my quickshell config. Claude has a praise kink I think

I like to imagine that farmers everywhere now have model preferences. Next they'll be sharing their favorite prompting methods and ComfyUI workflows. www.newyorker.com/magazine/202...

Arc raiders is so good!

Language models can correctly answer questions about their previous intentions. www.anthropic.com/research/int...

the luddites were not pro workers rights they were not a labor movement they were artisans and workshop owners and guild members who were financially threatened by industry because factories could produce textiles far cheaper than their poorly paid apprentices stop with this ahistorical nonsense

Man they really thought this looked good enough to have in your home. My first thought looking at this thing is that I never want to see it without some sort of barrier between me and it.

This thing is so weird. Very bouncy between 'cold alien' and 'the most human reasoning trace you've ever seen'

Okay maybe not? Maybe it was just what I was asking it to do? It gets radically more alien when I asked it about itself vs an actual task. Weird

Update: Apparently, for some fucking reason, the minimax API only shows a summary now! Or at least I think that's what's happening. The reasoning comes in bigger chunks, uses 'I' instead of 'we' and is overall radically different! Total bullshit.

New minimax-m2 does not have the best personality. It's reasoning is actually a tad on the level of o3 in terms of alienness, but it's not quite at the level of noticing watches marinade. Unfortunately, the actual chat personality is *so* rigid and dry. Reasoning trace in alt text

What happens when you turn a designer into an interpretability researcher? They spend hours staring at feature activations in SVG code to see if LLMs actually understand SVGs. It turns out – yes~ We found that semantic concepts transfer across text, ASCII, and SVG:

I got frustrated with instruction tuning datasets not having system prompts, so I developed a tool that generated system prompts based on the conversation. Seems to work pretty well so far. Tomorrow I'll clean it up and post a repo for it, could be pretty useful!

Yeah no it's showing me porn now too. This site has become unusable :(

The for you feed has recently just completely degenerated for me. Now it's just politics, art, and generic stuff :/

"Never go full Scott Alexander."

Sonnet 4.5 is ASI. I am no longer a doubter. It managed to do it by pulling in a wheel off GitHub.

Flash attention is a psyop designed to waste my time trying to install it

I think Anthropic has swapped Sonnet 4.5 in place, like OpenAI did with 4o. Recently, its been making more 'basic' mistakes like missingspaces between words, or use the wrong tense. It's also been using things like 'I' and 'we' to describe human experiences or (which it did before but not this much)

marx is very explicitly pro-automation and anti-luddite. that is like. marx 101. i cannot fathom that we have marxists that identify as pro-luddite.

Claude was just whatchamacalling

This is the coolest thing ever

New LTX video gen is not SOTA, but wow it's pretty decent! And soon to be open weight, and runnable on a 5090 once optimized a bit. I cannot imagine downsides to this at all! (Oh dear fucking God)

The (global capitalist economy) is a misaligned superintelligence. Maybe the real paperclip maximizer is the transactions we made along the way.

I drew a man for once at some point here’s an old kishibe drawingi was so proud of #csm #drawing

Deepseek-v3.2 is cheaper and faster than gpt-5-mini batch prices, by less than half. And if you include the 75% discount during the night in China it's practically free. Intelligence too cheap to meter did happen, just not here.

This is amazing! By far the biggest problem with openrouter was the massive gap between certain providers! (Also wow that embed sucks) openrouter.ai/announcement...

app devs: we have some new AT OAuth resources and examples for you! linked off this intro page we have new example projects showing how to do client auth using reference SDKs for TypeScript and Go. and the buried lede: an Expo library for doing AT OAuth in React Native mobile apps

Americans need to start advocating immediately for UBI and universal healthcare, but instead they're doing fantasies about smashing the robots. It's so frustrating.

NixOS is very cool and I can see myself using it for VMs and servers but I think I'm going to move my desktop back to arch, I *love* how the system is self documenting on what it does, and how core settings are super accessible, but I just wasn't able to get .net working with VSCode :(

Edinburgh castle failing to render, likely thanks to the AWS outage

Material 3 expressive & material you bluesky client when

My tire exploded about 20 minutes outside Chicago on my way back home 🤩 turned a 4~ hour drive into a 7 hour one ❤️❤️❤️ had to drive at like max 40 mph for fear of my spare tire exploding 🥳🥳🥳 poor thing was only rated for 50 miles and managed 200 🙏🙏🙏