Harlan Stewart, 17 April 2023
AI Impacts weblog
We moved our weblog to Substack! We predict this platform has many benefits, and we’re excited for the weblog to reside right here. Now you can simply subscribe to the weblog to obtain common newsletters in addition to numerous ideas and observations associated to AI.
AI Impacts wiki
All AI Impacts analysis pages now reside on the AI Impacts Wiki. The wiki goals to doc what we all know to this point about decision-relevant questions on the way forward for AI. Our pages have at all times been wiki-like: updatable reference pages organized by subject. We hope that making it an precise wiki will make it clearer to everybody what’s occurring, in addition to higher to make use of for this function, for each us and readers. We’re actively searching for methods to make the wiki even higher, and you may assist with this by sharing your ideas in our suggestions kind or within the feedback of this weblog publish!
New workplace
We just lately moved to a brand new workplace that we’re sharing with FAR AI and different companion organizations. We’re extraordinarily grateful to the workforce at FAR for organizing this workplace house, in addition to to the Lightcone workforce for internet hosting us over the past yr and a half.
Katja Grace talks about forecasting AI threat at EA World
At EA World Bay Space 2023, Katja gave a chat titled Will AI finish all the things? A information to guessing wherein she outlined a approach to roughly estimate the extent of AI threat.
AI Impacts within the Media
AI Technique
Resisted technological temptations: how a lot financial worth has been forgone for security and ethics in previous applied sciences?
Public consciousness and opinions about AI
- “The general public helps regulating AI for security” summarizes the outcomes from a survey of the American public about AI. (Zach)
- “How in style is ChatGPT?”: Half 1 seems at traits in AI-related search quantity, and Half 2 refutes a widespread declare concerning the progress of ChatGPT. (Harlan and Rick)
The state of AI at the moment: funding, {hardware}, and capabilities
Arguments for AI threat
Chaos principle and what it means for AI security
Miscellany
- “How unhealthy a future do ML researchers count on?” compares consultants’ solutions in 2016 and 2022 to the query “How optimistic or destructive will the impacts of high-level machine intelligence on humanity be in the long term?” (Katja)
- “We don’t commerce with ants” (crosspost) disputes the frequent declare that superior AI programs gained’t commerce with people for a similar purpose that people don’t commerce with ants. (Katja)
We’re actively looking for monetary assist to proceed our analysis and operations for the remainder of the yr. Earlier funding allowed us to develop our analysis workforce and maintain a summer season internship program.
If you wish to speak to us about why we ought to be funded or hear extra particulars about what we’d do with cash, please write to Elizabeth, Rick, or Katja at [firstname]@aiimpacts.org.
Should you’d wish to donate to AI Impacts, you are able to do so right here. (And we thanks!)
Picture credit score: Midjourney