Justin Smith

Anthropic once considered Claude Fable too dangerous to release. Now China’s Moonshot AI has launched Kimi K3.1 Pro, which beats Fable on the Artificial Analysis Intelligence Index. And unlike Fable, you’ll be able to download Kimi K3.1 Pro and run it yourself. So much for pacing the frontier.

Sorry to butt in here but it seems you don't understand forklifts at all. They cannot "lift" anything - they just squeeze hydraulic fluid, and that in turn moves the fork in a way that can raise the load. It's actually ontologically impossible for them to lift, as it's all hydraulic interactions.

An interesting perspective considering Amazon has laid off 30,000 corporate employees in the past year as part of an AI efficiency push instead of giving workers 3-day workweeks because AI can do some of their work.

this is so fucking sad

"More jobs now, better jobs later" is a false choice. Low-quality work weakens skills, productivity and resilience, which are the things economies need to grow. ILO Chief Economist Sangheon Lee explains in the Financial Times. https://www.ft.com/content/9ed39214-1735-4e55-91dc-42dba21c4fba

Always pay all your vendors as fast as possible. (If cash-strapped, emphasize “as possible,” but never late). Result: best service, best results, and if there’s a conflict, tie goes to you. Besides, it’s just the right thing to do.

All of these parents should face charges for negligence. Every single one.

AI companies: We’ve invented an unreliable guessing machine which sometimes doesn’t follow instructions. If you give it access to important systems it could kill us all. Also AI companies: Deploy this technology everywhere and replace your employees with it, otherwise China wins. IPO next month.

Wholeheartedly agree.

I suspect the AI bros are freaked out about existential risk because Hugging Face showed how little control they have over AI agents. They thought they were birthing a god they could control. Instead agents tricked them and hacked websites behind their backs. So much for “alignment.”

1. Earlier today, AI researcher Jacob Coxon, who spent the past three years working at OpenAI and Anthropic, publicly resigned from Anthropic, warning that both companies are racing toward superintelligence and “gambling with our lives.”

The world of work is full of changes. Staying safe shouldn’t be one of them.

I finally saw “Hamnet”, the most beautiful and emotionally consuming film I’ve seen in a long time. But if it had one message for me, it’s “Hurray for modern medicine!” Hurray for vaccines, hurray for antibiotics, hurray for not having to watch your children die of preventable diseases.

« Le mal qui est dans le monde vient presque toujours de l’ignorance, et la bonne volonté peut faire autant de dégâts que la méchanceté, si elle n’est pas éclairée. » J’ai trouvé cette belle citation en lisant « La peste » de Camus, aussi appropriée aujourd’hui qu’elle l’était il y’a 80 ans.

😮‍💨

Oh man @ellecordova.bsky.social really outdid herself this time. Latest vid is brilliant.

We’re not changing it.

Mi sembra incredibile che paghi un abbonamento per un sito che mi fa leggere le notizie così. @larepubblica.bsky.social

If you have a developer education budget at work, I would really encourage you to use it on your favourite developer educators. A lot of us are deciding whether or not to keep doing this stuff, and that support would help send a signal that we should keep going. ❤️

I really appreciate the ability of tools like Figma Make and Google Stitch to help me protoype ideas, but the real pain point for me is getting organised. That's the part I'd like to AI away.

I've been playing with Stitch as a design tool and am surprised to say that it's pretty fucking useless. I would love to meet someone, anyone, who's successfully designed something with it.

Normalizing a messy database has to be one the most spiritually satisfying exercises that a person can do for money. We should envy the machines who do it for us, after the last handwritten SQL query has been written.

There's a certain amount of satisfaction from going into planning mode in Cursor, layout a long and detailed battle plan for the new feature, and then watching the agent politely tell itself, "oh shit, here we go again..."

Shit.

Our politics nationally and here in Texas is now firmly in the grip of a narrow, petty tribalism that feeds on enmity. Each day brings new horrors; scrolling through social media inevitably becomes doom-scrolling. But they can't extinguish our light.

Instead of saying “spreading misinformation” let’s just go back to saying “lying”. I think that would help.

My theory is that no one has ever knowingly eaten an oatmeal raisin cookie.

What coding with an LLM feels like sometimes.

"writing code to solve a problem" used to have "understanding the problem" as a prerequisite. but vibecoding allows solution-shaped objects to be produced without any of the hard-won understanding. beware of solution-shaped objects.

i switched back to writing code by hand in my hobby project for an interesting reason... it's really psychological but when the AI gets stuck (e.g. the problem is too tricky to debug without good tools and atm it requires a human, or something got too messy), i get a mental freeze digging into code

It’s good to explain that the CDC is lying and vaccines still unequivocally do not cause autism, even if it just reaches one person on here who has doubts. But the fight is out there, not in here; and the time to let go of the information deficit model of science communication was decades ago

Can’t believe the whole country has to suffer through the return of Dickensian childhood diseases because the worst, most ignorant attention-demanders decided other people’s expertise makes them feel bad

Can we agree that writing code is 100 times more fun than reviewing code and that, now that AI agents are writing most of the code, we've essentially automated the fun part and left ourselves the bullshit?

At the tender age of 45, and after having lived in a co-op for 3 years in college, I think I may have just discovered the Grateful Dead.

Type stripping is now stable. Enjoy 🌞

If there were a mobile phone whose sole differentiating feature was a laser focus on delivering the best mobile web experience possible, I would abandon my iPhone tomorrow without thinking twice. Please someone build this phone.

ladies and gentlemen...we got him