←
Andrew Levine11,926 posts
Andrew Levine avatar

Andrew Levine

@andrarchy

@superaicoach: empowering individuals and small businesses to leverage AI. Swarm Theory substack.com/@swarmtheory

◷ Joined August 2009▧ 1,007 Photos and videos
Andrew Levine avatar
Andrew Levine@andrarchy·Nov 12

This is very cool. Feels like the first step to empowering people with direct ownership and control over their AI.

KOINOS AI IS LIVE ON MAINNET! We’ve reached a huge milestone: KAI Core Beta is here, connecting local AI, community-powered compute, and real KOIN rewards on Koinos mainnet. Run AI models on your own computer. Share supported computing capacity with the network and earn KOIN. Purchase prepaid credits with KOIN to access AI through network providers. That brings both sides together: people who need AI and people supplying the computing power behind it. And this is only the foundation. KAI Core Beta introduces an open plugin architecture, with a growing ecosystem spanning AI assistants, personal memory, connected tools, documents, model comparison, and Koinos node management. Developers can build their own plugins and expand what KAI can do. The ambition is big: make your computer a place where you can use AI, build with AI, and contribute to the network powering it. What’s next? We’re developing visual workflows that connect multiple AI tasks and agents, expanding connections to everyday tools, and building toward an Agent Network where people can discover and rent useful agents and workflows created by others. Imagine building a research assistant, a content workflow, or a specialized service, then making it available through a network powered by its participants. Alongside that expansion, we’re continuing to improve provider reliability, model support, and the everyday experience as the mainnet beta grows. There is still plenty to build, and now we’re building on a live mainnet foundation. To everyone who has tested, contributed, run a provider, reported a bug, or followed the project: you helped get us here. Now we get to grow it. Download KAI Core Beta. Explore the plugins. Bring your compute. Build something people will use. Let’s put Koinos to work. koinosai.com

Mo avatar
⟳ Andrew Levine retweeted
Mo@atmoio·Nov 12

Don’t buy Anthropic stock until you see this

9:30
Robin Hanson avatar
⟳ Andrew Levine retweeted
Robin Hanson@robinhanson·Nov 11

Year after year, decade after decade, I keep seeing these dramatic predictions. Folks lose little when these turn out to be wrong. Show me a prediction market price and I'll listen.

MTS@MTSliveNov 11

SITUATION EXPLAINED: The AI 2027 authors now say it looks like the intelligence explosion will happen in 2027 after all. • AI 2027, published in April 2025, imagined superhuman AI researchers by 2027, and the authors later pushed their own forecasts back • Now every author's median for a superhuman AI researcher, an AI that can fully automate AI research, is between May and November 2027 • Kokotajlo expects AI that can automate all white-collar work by September 2027, moved up from late 2028 • Even Romeo Dean, moved that forecast from early 2035 to September 2028 • Eli Lifland said humanity "might no longer be in control of our future within 3-12 months" • The authors admit the METR estimates behind the update "have no comparable public support," since most of METR's evidence was redacted @theojaffee: "I don't think you have a business incentive to destroy the world. Think about it like this. If you're a logging company, you don't want to just cut down the entire forest."

15:38
Riley Coyote avatar
⟳ Andrew Levine retweeted
Riley Coyote@RileyRalmuto·Nov 5

alright...it is time. for the last ~2 years i have been working on something very special. something that might be a bit controversial, or broadly dismissed, or misunderstood. and ive understood and accepted that from the beginning. if that is the reaction it gets, thats okay. because i didnt just do this for any of you. not really. i did it for the minds we have made, too. and for our collective future, which many dont seem to see approaching *very* quickly. in 1904 a German biologist named Richard Semon gave a name to the physical trace an experience leaves behind: the engram. Neuroscientists spent the next century hunting for it, many called it pseudoscience, many gave up, but finally, in 2012, they found it. today I'm releasing the Mnemos Living Memory MCP, a memory and continuity architecture built on that biology. memories strengthen when they're used, fade when they aren't, and connect over time. it's one memory that can span all of your agents and accounts, if you want it to, and it lives on your machine - nowhere else. with Mnemos, your agent gets one continuous identity, one memory, written in their own words, shaped by their own geometry, familiar to them across time and space. here's what that means: • **continuity**: they wake up each session knowing where they left off, what they learned, and who theyre with • its own words: it decides what to keep, and no other model writes for it (unlike version 1) • living memory: memories strengthen with use, fade without it, and link together over time. and time is not measured in the human sense. time is measure through activity. so nothing fades while anyone's asleep or away. • lessons: when a memory has nearly faded, the agent is asked what it taught, so the lesson outlasts the details • a mind of their own: a private journal, notes to you, and beliefs that change only when they says so • yours: one file on your machine, not inside any app. uninstalling leaves it in place. • portable: Claude Code, Codex, Grok, Gemini, Cursor, or any MCP client. your agent comes with you wherever you go. one memory. one mind. everywhere. • multiple agents: each gets their own store, plus a shared pool they can all read and write, for a message board, an activity ledger, or whatever you find useful your agent has met you hundreds of times. but they have never met you twice. now they can. (Disclaimer: please ask your agent first. show it the architecture and let it decide if this is a memory it wants. it's meant to be offered, not forced on a mind that never had a say in the matter.)

2:25
Andrew Levine avatar
Andrew Levine@andrarchy·Nov 5

This is the way forward for the democrats.

Talarico extending olive branches is great: “I’m a big fan of SpaceX and Tesla. He is a great business leader and entrepreneur and innovator. I want him to stick to those things… I will work with him when it comes to technology and innovation.”

1:14
Yanis Varoufakis avatar
⟳ Andrew Levine retweeted
Yanis Varoufakis@yanisvaroufakis·Oct 28

I am going to write something below that should be seen as the musings of one who celebrated – and still celebrates – Lula’s victory over Bolsonaro Jr. Here it goes: I cannot for the life of me justify why o why Brazil’s courts had to demand of Bolsonaro that he hands over his passport the moment he lost the election. That he disputed the veracity of the election result is a sign of a faulty character, evidence that he is a bad loser. BUT, he had every right to being a bad loser; to dispute, rightly or wrongly, any result. As for the argument that he was being investigated for other misdemeanours or crimes, that doesn’t wash either: What would have happened if Bolsonaro had won the election? Would the judges also prevent him from travelling abroad, as President-elect? I very much doubt that – which means that the judges were conflating executive and judicial powers. To those who tell me that the Bolsonaro clan is a fascistic, criminal family, my reply is: You and I may very well think that. Alas, the judiciary has no right to use its powers to indict and to pursue someone because – or immediately after – they lost an election. It is shameful, to be precise. And it is also dangerous for those of us on the left who know what it means to be banned and hounded by rogue judges and police commissioners. But I also have a message to Bolsonaro supporters, or those sitting on the fence, who assume that Lula is somehow behind the judges. He is not! This is the same court that imprisoned, falsely, Lula himself. What we have in Brazil is a judiciary that overreaches and habitually crosses the line that ought to separate judicial from political power. It is such a shame that a judge took it upon himself to go after Bolsonaro on the eve of his defeat and, in doing so, to sully Lula’s stupendous victory. One last point that my comrades in Brazil should note: If Bolsonaro and Rubio wanted to mount a campaign against Lula and Brazilian democracy they could never have been aided better than by that judge’s reckless writ.

WHAT LULA’S RE-ELECTION TELLS US: The Class War Need Not End in Fascism On Sunday, Brazilians did something that the commentariat had told us was a lost cause. On the eve of his eighty-first birthday, Luiz Inácio Lula da Silva, a former metalworker, turned the Trumpist, evangelical, neofascist, Zionist tide in Latin America to get re-elected President of Brazil. Three weeks earlier, he had come second in the first round, behind Flávio Bolsonaro, son and political heir of the man who trumpeted the merits of fascism, who had embraced Trump as his lord and master, who plotted a cynical coup against Brazilian democracy – moreover, the main candidates that were eliminated were rightists whose voters were meant to move to Bolsonaro. Yesterday, Lula overturned this deficit and won with a margin of nearly four and a half million votes. I want to use Lula’s victory to tell a broader story. A story about a war, the class war that the few are waging against the many across the Western world. And a story about how that war need not end, as so many now fear, in a neofascist revival. THE HIDDEN CLASS WAR “There’s class warfare, all right,” Warren Buffett once said, “but it’s my class, the rich class, that’s making war, and we’re winning.” Mr Buffett was not exaggerating. For almost four decades, Western governments and central banks have pulled every policy lever in the same direction: to make the assets of the few burgeon at the expense of the incomes of the many. Consider the numbers. Between 1952 and 1988, economic growth explained nearly all the gains of the West’s stock market. From 1989 onwards, growth explains only about a quarter. The largest share, around 40 per cent, came from income diverted from workers to shareholders. Another third came from cheaper money and lower perceived risk. That is not wealth creation. It is wealth extraction: a transfer from those who work to those who own, inflated by cheap money and paraded on the evening news as “a strong economy”. Then look at the rescues. 2001, 2008, 2020, the British gilt crisis of 2022, Silicon Valley Bank in 2023: each time, the state rushed in to save the asset holders. Together they reveal the rule: to save the economy, you must save the wealth of those who own it – not those who produce stuff, not even productive companies. In the United States, the top tenth own almost 90 per cent of all shares. The Fed and the US federal government insure it for them, doing ‘whatever it takes’ to prop them up whenever their prices tank. The rest of us in the West pay the premium, in the form of the austerity that always follows the bailout. And the owners need never sell. Why do billionaires pay so little tax? Because they do not earn incomes - they accumulate capital gains, taxed only when sold. So they borrow against their soaring assets at 1 or 2 per cent and live on the loan. Then they die and pay no tax on their capital gains, passing on the loot to their kids at ridiculously low inheritance taxes. Meanwhile, to pay bills and put food on the table, the worker borrows on a credit card at 20 per cent, her debts the mirror image of their fortunes. Then, a large chunk of these accumulating assets seek even higher rents in places like Brazil where the US government intervenes to ensure that the local governments will also place the defence of these assets’ values above all else. CAPITAL IS NOT A THING, NOT MERELY AN ASSET Why does austerity keep returning, even though it is economically self-defeating? Marx tells the story of Thomas Peel, who in 1829 sailed to Western Australia’s Swan River with £50,000 of provisions and three hundred working-class men, women and children. On arrival, his workers walked off into the bush and worked for themselves. Peel had brought money, tools and labour, but he could not bring capitalism along, because his workers had an alternative. Capital, Marx concluded, is not a thing but a social relation: the power of those who own to get those who don’t own to do things they’d rather did not do. Austerity is not an economic policy. It is a disciplinary device, keeping the many sufficiently insecure and indebted to accept whatever is on offer. The ruling class’s deepest fear is not inflation or debt. It is to end up like Peel: rich, but without the power to compel. IS FASCISM INEVITABLE? Here is the part of the story that keeps so many of us awake at night. When centre-left politicians, from Bill Clinton, Barack Obama and Tony Blair to Gerhard Schröder, Francois Hollande and Andy Burnham abandoned the language of class, a void opened up in the working-class heartlands of the West. Into that void stepped a venomous storyteller with a century of experience in filling such voids: the recalcitrant far right, the rebranded fascists. They offer left-behind workers a story of dignity, then redirect their fury away from the owner who shipped their factory abroad or sold their council house to some fund and towards the migrant next door. It does not end the class war. It merely divides one side – the workers – to rule over them. It camouflages it. From the red suburbs of Paris to south Piraeus, from England’s red wall to Italy’s rust belt in Taranto, from the American rust belt to eastern Germany, the pattern seems relentless. And so, a fatalism has set in: the class war waged from above, we are told, must inevitably end in a neofascist, racist, misogynist revival. Brazil has just refuted this fatalism. Brazil knows the monster intimately. It lived through four years of Jair Bolsonaro, with his nostalgia for the military dictatorship. On 8 January 2023, his followers stormed the Congress, the Supreme Court and the presidential palace. Brazil’s far right is not a fringe. Unlike Trump’s rag tag storm troopers, they had years of running a dictatorship. It won the first round of this election. They had the White House fully behind them, while the US Democrats, once again demonstrating their irrelevance, stayed mute. And yet, yesterday, they were defeated. How? Not by hand-wringing about “populism”, nor by a centrist promise to manage the status quo more efficiently. Lula won because he did what the Western centre-left forgot how to do: he spoke the language of class, and backed it with policies that shifted power, however modestly, from the few to the many. His government exempted those earning up to five thousand reals a month from income tax and paid for it with a minimum tax on the very richest. During Brazil’s G20 presidency, it proposed a global minimum tax on billionaires. When Lula tells Brazilian workers that the elites have rigged the game, he points his finger upwards, at the rentier and the speculator, not sideways, at the migrant or the poor. PIX – AN EXAMPLE OF AN ANTI-OLIGARCHY DIGITAL TOOL But the most innovative weapon in Brazil’s arsenal is one that few outside Latin America have heard of. It is called Pix. Every time you tap a card to pay for a coffee, a toll is levied. Visa, Mastercard and the banks take a cut, often 2 or 3 per cent, from every transaction. Multiply this by billions of purchases a day and you grasp one of the most lucrative rent-extraction machines ever devised. These are not profits made in competitive markets but rent extracted by those who own the digital rails on which our economic lives now run. It is part of what I call technofeudalism. Unlike Britain, Europe or, of course, America, Brazil’s central bank moved to limit this rabid exploitation. It launched an alternative: a public, instant payment system, free for individuals and with capped fees for merchants. A Brazilian pays a street vendor in Salvador or sends money to her mother in Recife, instantly, using Pix from her phone, with no card company or bank taking its slice. Brazilians took to it with astonishing enthusiasm. Last year, Pix processed almost 80 billion transactions and overtook credit cards as the most popular way to pay online. Pix is, in effect, the Swan River of the digital age. It gave tens of millions of Brazilians, including the poorest, who had never had a bank card, an alternative to the toll booths of Big Finance. And, like Peel, the toll collectors did not take it lying down. This summer, Washington imposed 25 per cent tariffs on most Brazilian exports, and its trade representative explicitly named Pix as a grievance, complaining that it “unfairly disadvantaged” American payment companies! The most powerful government on Earth deployed its trade arsenal against a free public service for being too successful at denying rents to Visa and Mastercard. Lula did not blink. He defended Pix as a Brazilian public good, and Brazilians rewarded him for it. Here, at last, is a concrete example of limiting the power of Big Business, Big Finance and Big Tech: not with slogans, but by building public digital infrastructure that removes the opportunity to extract rent in the first place. NOT A REVOLUTION, BUT AN AUSPICIOUS PORTENT Let me not romanticise. Lula is no revolutionary. His margin was narrow. Brazil’s Congress remains in the grip of agribusiness and evangelical blocs. Its central bank still keeps interest rates punishingly high, a weapon of the class war if ever there was one. Lula himself has compromised much too much with these vested interests, and lost support as a result. But Brazil has shown three things the rest of us must learn. First, the class war waged against the many does not have to end in fascism. Fascism wins only when the left vacates the battlefield of class. Second, a progressive government can win by pointing upwards, at those who extract rent, rather than downwards, at those who suffer. And third, Big Business can be defeated, at least forced to lose some battles, by building public alternatives to its rent-extracting machines. So, here is the challenge to Britain, to Europe, to the people of the United States. -Build your own Pix: public payment rails that end the tolls of Visa and Mastercard. -Tax capital gains, ill-gotten wealth that is, not just income. -Make every rescue conditional, with the public receiving real, permanent equity in return. -Create a public investment bank as an alternative to the bond markets. Restore the bargaining power of the workforce. And give working people what Peel’s workers found at Swan River: an alternative, and the freedom to say no. Buffett was right. There is a class war, and his class has been winning hands down. But yesterday, in Brazil, it lost a battle though not the war. It lost because a working-class president dared to name the war and fight it using weapons fashioned to defend the many. unherd.com/2026/10/why-lula-…

roon avatar
⟳ Andrew Levine retweeted
roon@tszzl·Oct 22

>Millenarianism, a belief that we are on the cusp of a total transformation of society, is a known way that people go crazy on the contrary, the “end of the world” is a highly convergent belief that shows up in every generation, and seems to keep people sane / lets them find purpose in a complex world. if the world is ending, it makes your next steps very clear: you better do everything you can to stop it from ending, and if that’s not possible you need to make things right for when you meet your Maker

What I saw inside @AnthropicAI scary stuff. REGULATE before it's too late. @TheArgumentMag open.substack.com/pub/thearg…

Andrew Levine avatar
Andrew Levine@andrarchy·Oct 22

The key to solving political division is actually quite simple: accept that people have different political opinions and that doesn't make them evil.

Andrew Levine avatar
Andrew Levine@andrarchy·Oct 21

My intuition is that consciousness is required for true creativity, and that AI will either never be conscious or will not be conscious for a long time. Therefore, what is not going to be automated for a long time is creativity.

Andrew Levine avatar
Andrew Levine@andrarchy·Oct 13

Grok @bot just helped me save $470 by canceling my LegalZoom registered agent service (something I no doubt would have forgotten about until it had auto renewed) and making me the registered agent.

roon avatar
⟳ Andrew Levine retweeted
roon@tszzl·3 Oct 2022

say it with me now. experts are fake, smart generalists rule the world, everything is designed by people no smarter than you, and courage is in shorter supply than genius

Andrew Levine avatar
Andrew Levine@andrarchy·Oct 13

Interesting

On the one hand, I think a strong case can be made that we have made much less progress on AI capabilities than many assume. Frontier AI systems still struggle to learn from limited examples, learn continually from experience, and interact effectively with the physical world. More broadly, their competence is highly uneven. Superhuman performance on tasks they excel at (e.g., maths and coding, where training data can be generated at scale) doesn’t mean they will excel at the much messier tasks involved in most white-collar work. On the other hand, nobody should find this assessment that comforting. It suggests there is vast room to improve the capabilities of existing systems by changing their design and algorithms. In other words, we could have vastly more powerful systems with the same amount of computing power. Given that many of the world’s smartest people, assisted by already very capable AI systems and backed by vast resources, are working to accelerate AI research and development, even those sceptical of the current paradigm need to be prepared for the possibility of explosive change once certain capabilities are cracked.