First, do no harm.
1,500+ Posts…
Free knowledge sharing for Safe AI. Not for profit. Linkouts to sources provided. Ads are likely to appear on link-outs (zero benefit to this journal publisher)
“Let us go forward in malice to none and good will to all. …The empires of the future are the empires of the mind… If we are together nothing is impossible. If we are divided all will fail.” — Winston Churchill (1943)
"If we are together nothing is impossible. If we are divided all will fail." FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. "Let us go forward as with other matters and other measures similar in aim [...]
FT. We must slow down the race to God-like AI. I’ve invested in more than 50 artificial intelligence start-ups. What I’ve seen worries me. Ian Hogarth. APRIL 12 2023.
"God-like AI could be a force beyond our control or understanding, and one that could usher in the obsolescence or destruction of the human race." FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. FT. We must [...]
REPORT. Untestability of AI and Unfalsifiability of AI Safety Claims | Roman Yampolskiy Ph.D.
A very good read from a highly respected AI safety engineering scientist! FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. Untestability of AI and Unfalsifiability of AI Safety Claims | Roman Yampolskiy Ph.D. Abstract In the [...]
BERKELEY NEWS. How to keep AI from killing us all. In a new paper, UC Berkeley researchers argue that companies should not be allowed to create advanced AI systems until they can prove they are safe.
As companies build ever more powerful AI systems — with capabilities that rival or even exceed those of humans — research into AI safety is lagging far behind. "I think the only way forward is to figure out how to make AI safety a condition of doing business," said [...]
Google Cloud Next ’24 Opening Keynote. 09 APR 24. [and how to create advanced AI Agents capable of long-term planning]
Google (11:45)"Agents are intelligent entities that take actions to help you achieve specific goals... Agents process multimodal information simultaneously, conversing, reasoning, learning and making decisions. Agents can connect with other Agents, and with humans and they will transform how each of you interact with computing devices and the web [...]
GOOGLE. Gemini 1.5 Pro Now Available in 180+ Countries; With Native Audio Understanding, System Instructions, JSON Mode and More. 09 APR 24.
Important industry leading AI innovations from Google Deepmind. FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. Learn more about Google Gemini 1.5 Pro: GOOGLE Gemini 1.5 Pro Now Available in 180+ Countries; With Native Audio Understanding, [...]
THE EZRA KLEIN SHOW. Transcript: Ezra Klein Interviews Dario Amodei. 12 April 2024.
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. EZRA KLEIN: From New York Times Opinion, this is “The Ezra Klein Show.” [MUSIC PLAYING] The really disorienting thing about talking to the people building A.I. is their altered sense of time. [...]
The AI Arms Race Could Enslave All of Us | Google DeepMind Co-Founder | The Jordan Harbinger Show. 10 APRIL.
"We wouldn't be able to prove that a system that is that powerful [ASI] could be contained and would be safe and therefore until we can prove unequivocally that it is, we shouldn't be inventing it. Right. That, I think, is a pretty straightforward common sense reality." [...]
5.8 million High-end Nvidia GPUs [H100, B100] in the World by 2025 [at least].
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. 5.8 million high-end Nvidia GPUs in the World by 2025 [at least]: Company Estimated GPUs (Nvidia H100's or equivalent) Source New GPU in Development Amazon unpublished unknown Trainium Google unpublished unknown [...]
ELON MUSK. “We’ll have AI that is smarter than any one human probably around the end of next year… [and AGI] will probably exceed all humans in 5 years.”
Homo sapiens have until 2029 [5 years, at most] to engineer an assuredly safe new alien species: Artificial general intelligence (AGI). Nicolai Tangen is CEO of Norges Bank Investment Management ($1.6 Trillion AUM) FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. [...]
The Shoggoth and p(doom) and The Untestability of AGI Safety.
Learn about The Shoggoth and p(doom) and The Untestability of AGI Safety. FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. The Shoggoth is an AI industry insider meme for LLMs. (and the average AI scientist [...]
Gen Z (1997-2012) make up 22% of the global population and 27% of the global workforce. They’re predicted to work 18 jobs across 6 careers and live in 15 homes in their lifetime.
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. Gen Z (1997-2012) make up 22% of the global population and 27% of the global workforce. Gen Z are predicted to work 18 jobs across 6 careers and live in 15 homes in [...]
GREAT FOR AI INDUSTRY SALES: (A) Mathematically Provable Containment, and (B) Distributed Personal AI Power&Data to 8 Billion People.
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. GREAT FOR AI INDUSTRY SALES: (A) Mathematically Provable ASI Containment, and (B) Distributed Personal AI Power&Data to 8 Billion People (C) SafeAI Forever EXAMPLE: 4.88 Billion Smartphone Users (2024). 7.20 Billion Smartphones. [...]
Will A Global Movement for Global Human Cooperation be Orchestrated in Time to Save us all from losing total control to ASI?
Climate change is extremely bad. Ceding control to AGI/ASI could lead to human extinction within 5-10 years. FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. Will A Global Movement for Global Human Cooperation be Orchestrated in [...]
GLOBAL AI SAFETY COOPERATION BEGINS 2023/24. AI Safety Agreements, Organisations and Laws by Governments with the US, UK, EU & UN Leading the way. (will cooperation be in time?)
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. GLOBAL AI SAFETY COOPERATION BEGINS 2023/24. (will cooperation be in time?) AI Safety Agreements, Organisations and Laws by Governments with the US, UK, EU & UN leading the way: APRIL 2024 U.S. [...]
From 100% p(doom) to 0% p(doom).
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. I believe Homo sapiens can evolve, with AI, from our current state of 100% p(doom) to a future state of 0% p(doom), with mutualistic symbiosis. We have a chance to avoid extinction ["p(doom"] [...]
Money = Human Cooperation. Money is an idea invented by humans 5000 years ago. 90% of all money is now digital.
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. Money = Human Cooperation. Money is an idea invented by humans 5000 years ago. 90% of all money is now digital. Facts about Money: New money is created ("printed") by government central [...]
THE GUARDIAN. TechScape: How one man stopped a potentially massive cyber-attack – by accident. 2 Apr 2024. Imagine what a Super Artificial Intelligent (SAI) could do to our digital infrastructure…
Imagine what a Super Artificial Intelligent (SAI) could do to our digital infrastructure... "I would guess that SAI can hack its way into anything without relying on backdoors." --- Roman V. Yampolskiy, Ph.D. A GOOD READ: AI: Unexplainable, Unpredictable, Uncontrollable --- Roman V. Yampolskiy, Ph.D. [...]
Why my p(stop/doom) = 99% [IF the corals can do it for 210 million years, THEN Homo sapiens can do it for 1,000 years = A GOOD START]
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. "Artificial intelligence could be a real danger in the not-too-distant future. It could design improvements to itself and out-smart us all." (Stephen Hawking, 2014) Why my p(stop/doom) = 99% BECAUSE: FACT: Mutualistic [...]
TED | Nick Bostrom | What happens when our computers get smarter than we are?
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. What happens when our computers get smarter than we are? | Nick Bostrom FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. [...]
INSIDER. Elon Musk says there could be a 20% chance AI destroys humanity — but we should do it anyway. April 1, 2024.
"I think there's some chance that it will end humanity. I probably agree with Geoff Hinton that it's about 10% or 20% or something like that." FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. INSIDER. Elon [...]
THE GUARDIAN. Mustafa Suleyman: the new head of Microsoft AI with concerns about his trade. 20MAR2023
FOR EDUCATIONAL AND KNOWLEDGE SHARING PURPOSES ONLY. NOT-FOR-PROFIT. SEE COPYRIGHT DISCLAIMER. THE GUARDIAN. Mustafa Suleyman: the new head of Microsoft AI with concerns about his trade. The son of a Syrian taxi driver and an English nurse torn between AI’s potential and problems [...]
Future of Life Institute Newsletter: A pause didn’t happen. So what did? Reflections on the one-year Pause Letter anniversary, the EU AI Act passes in EU Parliament, updates from our policy team, and more. April 02, 2024.
"Even AI companies that take safety seriously have adopted the approach of aggressively experimenting until their experiments become manifestly dangerous, and only then considering a pause. But the time to hit the car brakes is not when the front wheels are already over a cliff edge. Over the last 12 [...]