Source perplexity and citations JPMorgan Chase [ -4.68%] and Wells Fargo [ -1.34%] executives said at a Goldman Sachs [ +1.14%] conference on December 9 that AI is driving major productivity gains, with some operations seeing efficiency improvements of 40% to 50%, though they acknowledged workforce reductions are inevitable.[reuters +1]• JPMorgan’s Marianne Lake reported the bank doubled its productivity from 3% to 6% annually through AI, while Wells Fargo CEO Charlie Scharf said generative AI has made engineers 30% to 35% more efficient in writing code, and Wells Fargo’s workforce has already declined from 275,000 in 2019 to just over … Continue reading Major US banks say AI will boost productivity but cut jobs

Nvidia [-0.53%] CEO Jensen Huang issued a stark warning that China’s ability to rapidly build AI infrastructure could give it a decisive advantage over the United States, even as American companies maintain technological superiority in chip design.Speaking at the Center for Strategic and International Studies in late November, Huang highlighted a dramatic disparity in construction timelines. Building a data center in the United States “from breaking ground to standing up an AI supercomputer is probably about three years,” he said. By contrast, China “can build a hospital in a weekend,” underscoring the speed advantage that could prove critical as AI … Continue reading Nvidia CEO warns China’s infrastructure speed threatens U.S. AI lead

I found this fascinating insight from Anthropic. Image source: Reve / The Rundown ———- The Rundown: Anthropic researchers published a new study finding that Claude can sometimes notice when concepts are artificially planted in its processing and separate internal “thoughts” from what it reads, showing limited introspective capabilities. The details: Why it matters: This research shows AI may be developing some ability to monitor their own processing, which could make models more transparent by helping accurately explain reasoning. But it could also be a double-edged sword — with systems potentially learning to better conceal and selectively report their thoughts. Continue reading Claude shows limited ‘self-awareness’