forget the grind. small iterative steps. do things

Programmer/builder seeing world through lens of climate emergency, yet thrilled by the positive aspects of AI and general purpose technology @[email protected]

[forget the grind. small iterative steps. do things @halftroll]

countless lives could have been saved. Ignoring women in STEM/STEAM has consequences. Her contribution went well beyond topics like fresh air and ventilation. They included data collection—another area where WHO, CDC, and medical industry have been weak.

[Nathan Lambert @natolambert]

It's been a great effort by the early and growing American open-model labs since last June to put the US much more back on the map. We were getting totally owned last June.

Nvidia, Ai2, Arcee, Gemma, GPT-OSS and a few others will be seen as saving American open AI.

[Angelica 🌐⚛️🇹🇼🇨🇳🇺🇸 @AngelicaOung]

Jeremiah, I believed the Uyghur genocide thing during COVID too. And that the CCP is terrible an the Chinese people are all oppressed etc etc.

But the year is 2026 and Xinjiang has been open for tourism for years. Adrian Zenz is somehow the most authoritative source. I don’t think he’s even been to Xinjiang. His ideological bias is clear. He works for a group that’s anti communism. Would you like it if the world took the word of a Russian person working at an anti-American organization as the Chief Expert for how terrible America is?

Gaza is what a genocide looks like and despite the fact that Gaza is the size of Las Vegas and under Israeli lockdown we have a never ending stream of horrors from there on our timelines. What about Xinjiang, an area 1/6 of China total and crawling with tourists and good internet? Zenz and friends are still waving around those COVID era satellite images.

There is a terrible bug with liberalism. By assuming that your values are universal, you assume everybody who disagrees with you must be evil instead of just have different beliefs or priorities or maybe just because they know different things. You stop looking for pragmatic solutions, compromises, diplomacy and mutual understanding. You don’t even try to see the other persons point of view.

We’ve met each other. I’ve been on your podcast. I’m just the same Angelica you’ve had beers with in DC and Taipei. Then and now, I’ve always just wanted to make the world a better place. Isn’t it interesting how your instinct isn’t to be curious about why I changed my mind…it is to declare me beyond the pale? Aren’t you at all puzzled that a woman you previously thought so highly of your organization flew me out to DC to attend your conference suddenly decided to be a handmaiden to genocide?

No country or system is perfect and I’m no longer a naive believer in one-size-fits all. I hope that somehow you can get some objectivity about the Chinese system. I can’t think of another political party anywhere in the world that is as meritocratic, nor one that has consistently delivered such improvements in life for the last 3-4 decades to its people. Moreover I believe that Xi Jinping actually cares a lot more about the people of China than Donald Trump about ordinary Americans.

Just one more thing I want you to think about…I would almost certainly not have the beliefs I have today if not for living in Taiwan, talking to local Taiwanese, listening to their pundits and influencers. Like it or not, the Taiwanese are not all plucky-little China-hating boba-libs. If you really respect Taiwanese voices, you shouldn’t just cherry pick those who agree with you while dismissing everyone else as propaganda.

[Kevin S. Xu @kevinsxu]

American AI startups need open source models to survive and profit. They also don't care that they are made by Chinese labs, because if you control where you deploy the model and set up guardrails (aka you know what you are doing), the risk is minimal

This report by Fireworks AI and Harvey building legal AI agents using GLM, Kimi, and Opus is just one of many many examples.

No startup can afford to pay the Opus/GPT tax. Not sustainble.
https://t.co/YIc0xYsSFo

[Nathan Lambert @natolambert]

Nvidia joined the multi-teacher, on-policy distillation (MODP) gang! Is industry standard post-training right now.

The multi-teacher SFT to RL that Microsoft did in their first model was the standard established by DeepSeek R1. I expect MAI 2 to be MODP.

[Former Congresswoman Marjorie Taylor Greene🇺🇸 @FmrRepMTG]

The United States of America is a sovereign nation.

Section 224 of the 2027 National Defense Authorization Act must be removed.

Our military should not be integrated in any capacity with a foreign country’s military.

Nor should we be funding it.

[Nathan Lambert @natolambert]

I feel like this also goes for a lot of people without Mythos as they learn to use agents too tbf

[Thomas Piketty @PikettyWIL]

The world today is characterized by large-scale inequalities. And a climate crisis is looming over us.

We urgently need a new vision for global progress in the 21st Century. One that grounds human development and equality in planetary habitability.

What would it take to achieve high prosperity and equality while remaining within planetary boundaries?

The World Inequality Lab is very excited to launch the #GlobalJusticeReport.
[1/7]

[Bernie Sanders @BernieSanders]

Trey Martin is an ironworker, union president and eighth-generation Oklahoman who knows Oklahoma and the people who make it work.

I’m confident @treyforoklahoma will be the kind of representative Oklahoma working families can count on.

[Thomas Massie @RepThomasMassie]

On June 8, 2026, I’ll speak on the floor of the House to honor and memorialize the brave crew of the 🇺🇸 USS Liberty who died and were wounded in an unprovoked attack by 🇮🇱 Israel on June 8, 1967. Catch my speech on @cspan.

[forget the grind. small iterative steps. do things @halftroll]

@AngelicaOung This video shows something very different.

[Abubaker Abed @AbubakerAbedW]

Israel yesterday kidnapped four women. Two are footballers in the Palestinian National Team.

Their names are: Natali Abu Dia and Rand Halwani.

Is it normal to kidnap footballers, @FIFAcom? Where are sports media organisations? This story should be the headline everywhere.

[Richard Medhurst @richimedhurst]

Literally half the problems on this planet can be traced back to the UK. Including the US and Israel. It's extraordinary.

[Fei-Fei Li @drfeifei]

https://t.co/Kt50ttQRMJ

[AFRICAN & BLACK HISTORY @AfricanArchives]

On this day in 1972, Angela Davis was acquitted of charges stemming from a 1970 courtroom shootout.

Angela Davis, an iconic figure in the Civil Rights Movement and a prominent political activist, faced trial in 1972 on charges related to a courtroom shootout in Marin County, California.

Angela Davis, a scholar and member of the Communist Party USA, was a prominent advocate for prisoners' rights and an outspoken critic of racial injustice. In 1970, firearms registered to Davis were used in an armed takeover of a courtroom in Marin County, California, in an attempt to free the Soledad Brothers (3 black inmates charged with the murder of a prison guard).

The incident resulted in the deaths of four individuals, including a judge. Angela Davis was charged with aggravated kidnapping and first-degree murder in connection with the courtroom incident.

Her arrest and subsequent trial drew international attention, sparking a global "Free Angela Davis" campaign. The trial commenced in March 1972.

Davis was defended by a team of notable lawyers, including Howard Moore Jr., Leo Branton Jr. and Doris Brin Walker. The prosecution argued that Davis was complicit in the planning of the escape attempt, while the defense contended that she was being persecuted for her political beliefs and activism.

Her trial highlighted issues of racial inequality, the justice system's treatment of African-Americans, and political repression. Angela Davis was acquitted of all charges on June 4, 1972, after a 13-week trial. The jury concluded that she had not been involved in the planning of the kidnapping and murders.

[Millie Marconi @MillieMarconnni]

A psychologist who never built a computer wrote a paper in 1960 that described the personal computer, the internet, and AI assistants decades before they existed, then handed the money to the people who built them and let history forget his name.

I read about him at 1am. One name was missing from a story I thought I knew.

His name was J.C.R. Licklider. The book is The Dream Machine by Mitchell Waldrop.

In 1960, computers were room-sized machines that ran one job at a time. You wrote your program on punch cards, handed the stack to an operator, and waited days for your answer. Nobody touched the machine. Nobody talked to it. A computer on your desk that answered you in real time was science fiction.

Licklider was not a computer scientist. He was a psychologist who studied how the brain hears. But he used computers in his research, and one day he measured where his time went.

The result horrified him. 85% of his work hours were not spent thinking. They were spent getting ready to think. Plotting graphs by hand. Hunting for numbers. Reshaping one person's data to compare with another's. The insight took seconds. The setup took hours.

The problem was not that humans were slow. Humans and machines were doing the wrong jobs. Let the human ask the questions. Let the machine do the grunt work. Tie them so close they think as one.

He wrote it down in a paper called "Man-Computer Symbiosis." In it, a person sits at a screen and works with a computer in real time. The machine answers questions, runs the numbers, draws the results, pulls answers from everything it has seen. He was describing the laptop you are reading this on. He wrote it before most people had seen a computer.

A paper changes nothing on its own. Thousands of brilliant predictions die in a drawer.

What made Licklider different is what he did next.

In 1962, the Pentagon put him in charge of a research office at ARPA. He had a budget and near total freedom over where the money went. Most people would have funded the safe things. He did the opposite. He spent government money on a dream with no military use and no promise it would work.

He found the few researchers across the country who thought like him. He gave them money. Real money. No strings. He funded the work that became time-sharing, the first computers people could talk to. He funded the labs that built the mouse, the window, the screen. He built computer science departments where none existed.

He was not picking projects. He was building a tribe.

Then came the idea that should make you stop. In 1963, he sent a memo to everyone he funded. He addressed it, half joking, to the "Members and Affiliates of the Intergalactic Computer Network." Inside, he asked a question nobody else was asking. What if all these separate computers could link together, so anyone could share information and build on each other's work?

He was describing the internet. No network existed yet. He sketched it thirty years before it reached your house.

He left in 1964. He never built the network himself. But the men he funded carried it forward. His successors took his memo and turned it into ARPANET, the first working internet, a few years later. The researchers he paid built the personal computer at a lab called Xerox PARC. Every piece of the world he imagined got built by the people he gathered and funded.

Here is the part I cannot shake.

He gave away the credit on purpose. He did not want his name on the breakthroughs. He believed the vision had to outlive him, so he made the people around him strong enough to carry it without him. He won so completely that the vision survived and the man vanished.

Ask who invented the internet and you will hear a dozen names. Almost none will be his. The man who saw it first, wrote it down, and paid for it, is a footnote in the story he started.

He died in 1990. He never owned a personal computer that worked the way he dreamed. He never browsed the web. He never saw the thing he funded swallow the planet.

Every screen you talk to today runs on an idea one quiet psychologist had while staring at how much of his life was wasted not thinking.

He did not want the credit. He wanted the future.

He got the future. We just forgot who paid for it.

[Giada Pistilli @GiadaPistilli]

For months, I've been deep in the weeds of what it takes to make LLMs work well across languages, and I'm thrilled that this work now has a dedicated home: I'm leading a newly created multilingual team within Science at @MistralAI 🌍

[forget the grind. small iterative steps. do things @halftroll]

@SahilBloom Do cheese/yogurt count as single ingredient foods? Lemon on fish too, I hope.

[Pedro Sánchez @sanchezcastejon]

En la noche de ayer, el conflicto en Líbano se saldó con cuatro ataques de mortero a la base española Miguel de Cervantes, un casco azul serbio fallecido y dos españoles heridos.

Nuestra condena más absoluta a la violencia y todo nuestro apoyo a quienes se juegan la vida por defender la paz bajo la bandera de las Naciones Unidas.

Esperamos que todas las partes cumplan íntegramente el alto al fuego anunciado ayer y que las hostilidades terminen. La paz es el único futuro.

[Dwarkesh Patel @dwarkesh_sp]

Recently met @srush_nlp and he started giving me an impromptu lecture on how targeted on-policy self-distillation works.

I asked him if I could record it on my iPhone.

The basic idea is this: if the model made a mistake at some point in the rollout (for example, calling a tool that doesn't exist), we want to discourage this specific error, but we don't want to just learn from the final reward, because it's a very noisy signal spread out over the whole trajectory.

So we have another model read this trajectory and figure where the error was made. It simply inserts some hint tokens to the part of the trajectory right above where the mistake was made.

Now with these injected hint tokens, have the model run a forward pass. You're not having to regenerate a new rollout - aka no new decode required.

The hint causes the model to assign lower probabilities to the error tokens. You then trains the original model to match these new probabilities, teaching it to downweight that specific mistake.

[Amy Klobuchar @amyklobuchar]

Good — now let's pass it in the Senate and get this done.