Post AScRFCk0JPIpLyLS5Y by bkeegan@hci.social
 (DIR) More posts by bkeegan@hci.social
 (DIR) Post #AScRFCk0JPIpLyLS5Y by bkeegan@hci.social
       2023-02-09T04:21:20Z
       
       1 likes, 0 repeats
       
       You can draw a direct line between Alphabet losing $100 billion in market cap for over-promising on its AI efforts and its dismissing world-class researchers like @timnitGebru, @meg, and @alex who were pumping the brakes on this exact type of AI hype years ago.
       
 (DIR) Post #AScRFDvjtGHf2eqL1E by bkeegan@hci.social
       2023-02-09T04:31:06Z
       
       0 likes, 0 repeats
       
       Google literally had a more than two-year head start on issues of ethics, fairness, and trust in LLMs. If I was a $GOOGL stockholder, I’d have some pointed questions about leadership prioritizing the feelings of engineering managers over the entirely justified concerns about strategy and implementation raised by its expert researchers.
       
 (DIR) Post #AScRFEq6VhN1rTDLsG by bkeegan@hci.social
       2023-02-09T04:57:13Z
       
       0 likes, 0 repeats
       
       You have to go all the back to, say, February 2022 to find an analogous mania for a technology with fundamental technical gaps and massive social liabilities. And all those golden boys were bankrupt or indicted just 10 months later. The real winner of this AI race will be who can scale LLM training and content moderation by repackaging the most human labor at the lowest cost into an inoffensive simulation of a computer.