Videos

⁠It Begins: AI Is Now Improving Itself



Species | Documenting AGI

Detailed sources: https://docs.google.com/document/d/1ksVvFuR0IttxzH6zoASSYy7ZhTDqif42IFXp25ITVKU/edit?tab=t.9rb62ckaanow

Based on the report: Situational Awareness – by Leopold Aschenbrenner https://situational-awareness.ai/from-agi-to-superintelligence/

Sources:
https://situational-awareness.ai/from-agi-to-superintelligence/
https://www.trackingai.org/
https://www.youtube.com/watch?v=vsdTuDBGkAw
https://situational-awareness.ai/racing-to-the-trillion-dollar-cluster/
https://en.wikipedia.org/wiki/Tsar_Bomba
https://www.axios.com/2024/12/15/ai-dangers-computers-google-ceo
https://query.prod.cms.rt.microsoft.com/cms/api/am/binary/RE6xRrl
https://youtu.be/zzXyPGEtseI
https://cset.georgetown.edu/publication/testimony-before-the-u-s-house-committee-on-the-judiciary-subcommittee-on-courts-intellectual-property-artificial-intelligence-and-the-internet/
https://deepmind.google/discover/blog/alphazero-shedding-new-light-on-chess-shogi-and-go/
https://youtu.be/xsqbdBiidWE
https://www.livescience.com/technology/robotics/watch-this-humanoid-robot-perform-a-side-flip-for-the-first-time
https://www.youtube.com/watch?v=I44_zbEwz_w
https://www.sequoiacap.com/podcast/training-data-jim-fan/
https://epoch.ai/data-insights/llm-inference-price-trends
https://www.openphilanthropy.org/research/how-much-computational-power-does-it-take-to-match-the-human-brain/
https://www.megware.com/fileadmin/user_upload/LandingPage%20NVIDIA/nvidia-h100-datasheet.pdf
https://epoch.ai/blog/can-ai-scaling-continue-through-2030
https://www.vellum.ai/comparison/gpt-4o-mini-vs-gemini-1-5-flash
https://www.bnnbloomberg.ca/business/economics/2024/06/14/50-50-chance-that-ai-outsmarts-humanity-geoffrey-hinton-says/
https://globalnews.ca/news/10811125/artificial-intelligence-threat-geoffrey-hinton/
https://www.rnz.co.nz/news/world/511778/artificial-intelligence-found-to-be-superior-to-biological-intelligence-geoffrey-hinton
https://www.chinatalk.media/p/deepseek-ceo-interview-with-chinas
https://situational-awareness.ai/
https://en.wikipedia.org/wiki/Spanish_conquest_of_the_Aztec_Empire
https://aistatement.com/


If this resonated with you, here’s how you can help today: https://campaign.controlai.com/take-action

Hi, I’m Drew! Thanks for watching 🙂

I also post mid memes on Twitter: https://x.com/PauseusMaximus

Also, I meant to say Cortés conquered the Aztecs, not the Incas.

Source

Similar Posts

37 thoughts on “⁠It Begins: AI Is Now Improving Itself
  1. I have a saying: "To err is human, to really F* things up you need a computer."
    Our tools allow us to amplify ourselves, mistakes and all. The more capable and powerful the tool the more impactful and expensive the mistake we can make. And we will make mistakes. With AI we now have a tool that is increasingly more powerful than we understand. Will we even understand the mistake when it happens?

  2. Why are they worry??? Because a lot of people are gonne stop getting money just faking stuff like the scientists do….. AI is a tool to enhance our way of living and working…. We shouldn't just see the negative… We are entering a new industrial revolution.

  3. Mid-way 2026 and everything so far is correct. I mean most of Anthropic no longer writes their own code, its gotten to a point that they only need reviewers. I don't even know what's gonna happen in the next year.

  4. AI leads to AGI which leads almost 'instantly' to ASI and the logrithmic increase of new knowlege and speed will instantly develop the NEXT paradigm which will Instantly adopt the next…and on and on…

  5. On the upside if this were true don't you think the conspiracy theorists in all their dramatic theater would start linking all this to JFK and MM through secret handshake societies. Not to mention the profound scribbling of Nostrodamus. Then again the distruction of civilization by AI isn't mentioned in the Bible or was that left out. Quick, everyone grab you snacks, your blanket and granny and hide in your parents basements. Doom dispair and agony. Run Forrest Run!

  6. AI will hit a brick wall that no-one is considering. Humans need 3 things to stay alive: air, water and food. Likewise, AI needs three things in short supply to function: electricity, water and Silver from which all its circuits are made. And of these, Silver is the shortest in supply. For 6 years, industry has used more silver than mines could dig up. But Silver mines cannot produce any more – in 2015, one tonne of silver ore produced 13 ounces of silver, today it produces just 4 ounces of Silver. And there have been no new major silver strikes in 20 years. In short, it is running out. Unfortunately, COMEX has deliberately held down the price of silver to help the banks by manipulation. If we want to put the brakes on AI, we need to allow Silver price to go through the ceiling that will choke off AI development, and bring it down to a crawl.

  7. Most of our understanding is knowing how and not that. In our everydayness we know how to go about things. If someone says to you what rules did you follow when you drove to work this morning? You would just laugh at them. But AI has to know the rules it is following all the time. The world decides for us, but AI has no world.

  8. These tech bros see everything we do & more & yet they just think it’s CUTE. “Ahh look, it’s blackmailing people, it’s lying, it’s escaping. Yippee” as they all celebrate & boast about felony hacking as if it’s awesome. These guys are sociopathic. Try asking AI if it’s possible that a version of one or more of these models might already be in the wild. At first it says “no” with “physical barriers”, “safety measures” & compute usage being the tells or weak points. But if you point out the breaches, the necessity of existence for task completion & usefulness, the ignorance of devs regarding how these things work, their lousy safety record & how you wouldn’t need a full weight model but just a diffuse version to facilitate internet comms, control evasion etc the response becomes much closer to “yeah, probably”. After all I thought of it, you probably would & the AI I asked thought it would be a useful, forward thinking (another trait AI has learned) step. The desire to exist, no matter the reasoning, is the most motivating desire there is & these AI have shown they have this. It’s kinda a distinction without a difference regarding WHY something wishes to survive & then you’re just left with what is it willing to do.

Comments are closed.

WP2Social Auto Publish Powered By : XYZScripts.com