Victor Li

12 followers · 10 following · 1060 views

on the atlas — 33

highlights — 159

Even if one’s original plan cannot possibly work, a willingness to pivot and expand and notice failure or inadequacy means you are still in the game.
OpenAI Launches Superalignment Taskforce
Flaws in the alignment and understanding of each AGI likely get carried over in sequence. If you use A to align B to align C to align D, that is at best a large game of telephone. At each stage the one doing the aligning is at a disadvantage and things by default only get worse.
OpenAI Launches Superalignment Taskforce
If the first stage in your plan for alignment of superintelligence involves building a general intelligence (AGI), what makes you think you’ll be able to align that first AGI?
OpenAI Launches Superalignment Taskforce
how can a concrete, physical system perform a computation when computation is defined by an abstract mathematical formalism?
Computation in Physical Systems (Stanford Encyclopedia of Philosophy)
Again, what does the rock lack that a genuine computer has?
Computation in Physical Systems (Stanford Encyclopedia of Philosophy)
If we want to make progress in biology, we need a high-status way for top biology researchers to keep doing research, in the same way that DeepMind, OpenAI, FAIR, etc. have created a new high-status way for AI researchers to keep doing research.
Why is progress in biology so slow?
How much progress we can make in biology in the next hundred years will depend on the extent to which the language models are able to solve these problems.
Why is progress in biology so slow?
The best idea I can see at the moment is to use machine learning to predict drug safety, and then work with regulators to enable drugs that are predicted with high confidence to be safe, bioavailable, and to have good PK/PD etc. to be tested directly in human patients without other preclinical trials.
Why is progress in biology so slow?
One simply needs to have a sufficiently deep understanding of the structure and dynamics of the focal system, and the ability to monitor it. This is the domain of the dragon kings.
Dragon king theory
Taleb claims that black swan events are not predictable, and in practice, the theory encourages one to "prepare rather than predict", and limit one's exposure to extreme fluctuations.
Dragon king theory
“It’s one thing to have stories on the conspiracy blogs, but this takes it to the next level, with genuine insiders coming forward,” Pope said. “When these people make these formal complaints, they do so on the understanding that if they’ve knowingly made a false statement, they are liable to a fairly hefty fine, and/or prison. “People say: ‘Oh, people make up stories all the time.’ But I think it’s very different to go before Congress and go to the intelligence community inspector general and do that. Because there will be consequences if it emerges that this is not true.”
US urged to reveal UFO evidence after claim that it has intact alien vehicles | UFOs | The Guardian
“I hope this revelation serves as an ontological shock sociologically and provides a generally uniting issue for nations of the world to re-assess their priorities,” Grusch said.
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
Dr. Garry Nolan, a Professor in the Department of Pathology at Stanford University and a renowned inventor and entrepreneur with more three hundred published papers, has started over half a dozen companies based on technologies out of his laboratory. Nolan has previously applied some of those technologies to the analysis of exotic materials, publishing the first peer-reviewed paper examining such materials.
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
Grusch left the government on April 7, 2023, in order, he said, to advance government accountability through public awareness. He remains well-supported within intelligence circles, and numerous sources have vouched for his credibility.
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
“The non-human intelligence phenomenon is real. We are not alone,” Grey said. “Retrievals of this kind are not limited to the United States. This is a global phenomenon, and yet a global solution continues to elude us.”
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
Analysis has determined that the objects retrieved are “of exotic origin (non-human intelligence, whether extraterrestrial or unknown origin) based on the vehicle morphologies and material science testing and the possession of unique atomic arrangements and radiological signatures,” he said.
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
Other intelligence officials, both active and retired, with knowledge of these programs through their work in various agencies, have independently provided similar, corroborating information, both on and off the record.
Intelligence Officials Say U.S. Has Retrieved Craft of Non-Human Origin - The Debrief
but even then there's no guarantee that all sightings will be explained.”
Mysterious Metallic Orbs Flying ‘All Over the World,' Pentagon Says at NASA UFO Panel
Only about 2 to 5 percent of reported UAP sightings turn out to be “really anomalous,” Kirkpatrick said.
Mysterious Metallic Orbs Flying ‘All Over the World,' Pentagon Says at NASA UFO Panel
We see these all over the world and we see these making very interesting apparent maneuvers.
Mysterious Metallic Orbs Flying ‘All Over the World,' Pentagon Says at NASA UFO Panel
NASA’s UAP study team was convened in 2022 with the mission of investigating the origin and nature of UAPs with rigorous scientific standards using mostly unclassified data.
Mysterious Metallic Orbs Flying ‘All Over the World,' Pentagon Says at NASA UFO Panel
Hong claimed to have experienced mystical visions after failing the imperial examination.
Hong Xiuquan
When we put seven pages of reporting findings and questions to Worldcoin, the company’s response was that nearly everything negative that we uncovered were simply “isolated incident[s]” that ultimately wouldn’t matter anyway, because the next (public) iteration would be better.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Stronger data protection laws in Europe and the United States mean that the most ambitious entrepreneurs in those regions can’t get all the training data that they need from their own populations, she says, so they have to look to the developing world.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
When he initially expressed similar concerns, he was assured by a representative that all his questions were addressed in the Worldcoin “white paper.” No such document exists. According to the company, this is by design—people would be unlikely to read “a long, highly technical academic-style paper,” it said, and its shorter blog posts could be thought of as white papers.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
He wishes that he had received “a brief training or basics about Worldcoin.” Instead the only instruction he got was to “bring more people in to get yourself more money,” he said. “The rest was up to my social marketing skills.”
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Worldcoin has promised to open source its code, including repeating to MIT Technology Review on multiple occasions that this would occur “within the next few weeks”—since we first contacted the company in February.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
why did Worldcoin target lower-income communities in the first place, instead of crypto enthusiasts or communities?
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Additionally, its data consent form notes that the company also conduct “contactless doppler radar detection of your heartbeat, breathing, and other vital signs.” In response to our questions, Worldcoin said it never implemented vital sign detection techniques, and that it will remove this language from its data consent form. (As of press time, the language remains.)
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Additionally, by biometrically confirming that an individual is human, Worldcoin would solve another “very fundamental problem” in decentralized technologies, according to Blania: the risk of so-called Sybil attacks, which occur when one entity in a network creates and controls multiple fake accounts.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Our investigation revealed wide gaps between Worldcoin’s public messaging, which focused on protecting privacy, and what users experienced. We found that the company’s representatives used deceptive marketing practices, collected more personal data than it acknowledged, and failed to obtain meaningful informed consent.
Deception, exploited workers, and free cash: How Worldcoin recruited its first half a million test users | MIT Technology Review
Cannibalism has been said to test the bounds of cultural relativism because it challenges anthropologists "to define what is or is not beyond the pale of acceptable human behavior".
Human cannibalism
“Oh, and you can quote me on that last bit.”
What Really Made Geoffrey Hinton Into an AI Doomer | WIRED
“They just had to make something go bang, but it is a lot harder to make sure something doesn’t,” he says.
What Really Made Geoffrey Hinton Into an AI Doomer | WIRED
“I listened to him thinking he was going to be crazy. I don't think he's crazy at all,” Hinton says.
What Really Made Geoffrey Hinton Into an AI Doomer | WIRED
Another issue is that Demis Hassabis is not going to be at the meeting. A bad sign for his influence over how things play out, and therefore a bad sign for alignment and humanity, given the alternatives. Or could be a logistic issue, but for this, you make the trip.
AI #10: Code Interpreter and Geoff Hinton
Pro tip: If you see yourself saying this in the future, consider that you might want to stop whatever it is you are doing.
AI #10: Code Interpreter and Geoff Hinton
Jeffery Ladish is worried about the Agency Overhang problem, where we have LLMs that are superhuman in some ways while missing other capabilities, most importantly agency. What happens if you managed to give them the kind of agency humans have?
AI #10: Code Interpreter and Geoff Hinton
We have a huge ‘job overhang’ and there will be plenty we want humans to do once they are free to do other things and we are wealthy enough to hire them for currently undone tasks.
AI #10: Code Interpreter and Geoff Hinton
For example, I had forgotten or missed Altman’s trick of substituting the word software for the word AI. This does seem useful, while also serving to disguise many of the ways in which one might want to be concerned.
AI #10: Code Interpreter and Geoff Hinton
DeepMind releases work on robot soccer. If your model of the future involves ‘robotics is hard, the AI won’t be able to build good robots’ then decide for yourself now what your fire alarm would be for robotics.
AI #10: Code Interpreter and Geoff Hinton
When people say ‘there is no secret’ one should always suspect this means ‘there is no secret I can easily explain to you or share with you’ or more simply ‘you have to do the work, and you won’t get a big return from marginal efforts even if they’re clever.’
AI #10: Code Interpreter and Geoff Hinton
In general, the right answer on the margin is to choose to learn such things.
AI #10: Code Interpreter and Geoff Hinton
Will AI take lawyers’ jobs, or will we simply sue each other more? Will it take designers’ jobs, or will we demand better design? Will it take doctors’ jobs, or will we all demand our own concierge doctors? Will it take engineers’ jobs, or will we demand more personalized and higher-quality software? Will it replace startups, or will we demand more diverse and better instantiations of founders’ visions, faster? Will it take writers’ jobs, or will readers just demand higher-quality, more well-researched, more original work?
Intelligence Superabundance - Not Boring by Packy McCormick
One thing I love about customers is that they are divinely discontent. Their expectations are never static – they go up. It’s human nature.
Intelligence Superabundance - Not Boring by Packy McCormick
AI and more human lawyers
Intelligence Superabundance - Not Boring by Packy McCormick
when something becomes more efficient, people consume more of it.
Intelligence Superabundance - Not Boring by Packy McCormick
“What would I do with superabundant intelligence?”
Intelligence Superabundance - Not Boring by Packy McCormick
history suggests that we’ll find ways to put it all to work, with latent demand left to spare.
Intelligence Superabundance - Not Boring by Packy McCormick
“When you see something that is technically sweet, you go ahead and do it.” He does not say that anymore.
‘The Godfather of A.I.’ Quits Google and Warns of Danger Ahead - The New York Times