“Today, we launch YouAgent, an AI agent with code execution for more accurate answers to complex math and science questions. Unlike other consumer chat engines that cannot do reliable, multiple-step reasoning, it will also compute your mortgage accurately...YouAgent enables the LLM to write Python code, run that code and then use the output from the executed code to answer complex STEM questions more accurately than any other pure LLM chatbot on the market.” https://about.you.com/introducing-youagent-with-code-execution/
Microsoft announced Copilot, your everyday AI companion https://blogs.microsoft.com/blog/2023/09/21/announcing-microsoft-copilot-your-everyday-ai-companion/
How AI agents are already simulating human civilization https://venturebeat.com/business/how-ai-agents-are-already-simulating-human-civilization/ [https://archive.ph/wip/2D8mP]
TRI’s Robots Learn New Manipulation Skills in an Afternoon. Here’s How. https://medium.com/toyotaresearch/tris-robots-learn-new-skills-in-an-afternoon-here-s-how-2c30b1a8c573 [https://archive.ph/wip/aVyJ2]
"We don't let individuals or corporations develop nuclear weapons for private use, and we shouldn't allow this to happen with dangerous, powerful AI." https://time.com/6314045/prevent-ai-disaster-nuclear-catastrophe/
"54% of the 2,063 adults … survey[ed] in July said they were more concerned about the risks of AI than they were excited about the potential benefits … 39% …said they believed today's AI technologies are safe & secure — down 9 points from … November 2022." https://www.axios.com/2023/09/19/mitre-harris-poll-ai-security-risks
“In the new AI Policy Institute/YouGov poll, the “better us than China” argument was presented five different ways in five different questions. Strikingly, each time, the majority of respondents rejected the argument. For example, 67 percent of voters said we should restrict how powerful AI models can become, even though that risks making American companies fall behind China. Only 14 percent disagreed.” https://www.vox.com/future-perfect/2023/9/19/23879648/americans-artificial-general-intelligence-ai-policy-poll
Michael Nielsen's "Notes on Existential Risk from Artificial Superintelligence" https://www.lesswrong.com/posts/8Q7JwFyC8hqYYmCkC/link-post-michael-nielsen-s-notes-on-existential-risk-from
“When asked by the audience what exactly these recommendations to address risk from AI would look like, I recommended: ... 3. A Global AI ‘Kill Switch’ - where governments build the infrastructure to be able to shut down powerful AI systems deployments and training runs if needed, and regularly test it with fire drills.” https://twitter.com/NPCollapse/status/1704904095155245194
Liquid Computer Made From DNA Comprises Billions of Circuits https://www.sciencealert.com/liquid-computer-made-from-dna-comprises-billions-of-circuits
Powerful, soft combustion actuators for insect-scale robots https://www.science.org/doi/abs/10.1126/science.adg5067
Human trials of artificial wombs could start soon. Here’s what you need to know https://www.nature.com/articles/d41586-023-02901-1 [https://archive.ph/8fyNo]
Is the future of energy ... pouring water on hot rocks in the ground? Geothermal’s “breakthrough,” and the challenges ahead, explained. https://www.vox.com/future-perfect/23825844/geothermal-enhanced-fervo-demonstration-superhot
Ukraine links:
“Ukrainian forces have now gotten heavy equipment past the fighting positions - the third layer within the trilayered belt of field fortifications.” https://twitter.com/TheStudyofWar/status/1705041912481423825
AFU has fire control over the Bakhmut-Horlivka highway. https://twitter.com/NOELreports/status/1704809973128085805
Aftermath of the Ukrainian strike at the Black Sea fleet command post near Sevastopol. https://twitter.com/kromark/status/1704801675087630637
“Ukrainian forces hit a Russian base in Melitopol at a local motor factory. As a result of the attack, the commander of the 58th Army of the Russian Federation, as well as their chief of staff, were wounded. In addition, several dozen Russian officers were injured.” https://twitter.com/NOELreports/status/1704832343268118980
A large-scale fire strike was carried out at the Saki military airfield in the temporarily occupied Crimea https://twitter.com/front_ukrainian/status/1704752289821397147
They strive for “maximum authenticity” while producing the fakes as cheaply as possible. Everything lying around the factory is used: plastic tubes, foam blocks, used tires, scrap wood & metal. Decoy radar reflectors are made from cut-up Lukoil barrels. https://www.ft.com/content/b0581f55-a449-439c-a92f-1dfb1ca5a181 [https://archive.ph/okNSN]
YouAgent's code execution capability sounds very cool, but how does this differ from ChatGPT Pro's code execution capabilities? I can already do sophisticated data analysis/Python scripting in ChatGPT. Here's ChatGPT calculating the mortgage question YouAgent uses in its example: https://chat.openai.com/share/ced5922b-225a-473e-95b7-2ab14163df48
EDIT: poking around on You.com's web site I find this more robust demonstration of what they're talking about, which does seem impressive. https://you.com/search?q=%40agent+Approximately+3.5+percent+of+all+children+born+in+a+certain+region+are+from+multiple+births+%28that+is%2C+twins%2C+triplets%2C+etc.%29.+Of+the+children+born+in+the+region+who+are+from+multiple+births%2C+22+percent+are+left-handed.+Of+the+children+born+in+the+region+who+are+from+single+births%2C+11+percent+are+left-handed.%0A%28a%29+What+is+the+probability+that+a+randomly+selected+child+born+in+the+region+is+left-handed%3F%0A%28b%29+What+is+the+probability+that+a+randomly+selected+child+born+in+the+region+is+a+child+from+a+multiple+birth%2C+given+that+the+child+selected+is+left-handed%3F%0A%28c%29+A+random+sample+of+20+children+born+in+the+region+will+be+selected.+What+is+the+probability+that+the+sample+will+have+at+least+3+children+who+are+left-handed%3F%0A&cid=c1_a165e100-c3d7-42d0-b6b1-1a7c6c64be98&tbm=youchat