Welcome to the Sunday edition. This week:
Tech on the ballot: On Dario Amodeiās beliefs and why Americans hate AI
Is AIās productivity payoff finally arriving? One strong US quarter isnāt proof.
Europe is financing the tech race itās losing.
Upgrade to enjoy all EV perks ā some of the best things happen between our editions:
A walk with Azeem, beta-testing our MCP, and live member conversations youāll want to join.
Tech on the ballot
Tech wasnāt formally on the ballot this week, but it was in reality. I was struck by this comment by Gina Hinojosa, the Governor-Elect of Texas: āI say that the left-right paradigm just does not meet the moment of our time. What this is about is about 99.9% of us in the same boat against like 15 billionaire tech guys who are controlling governmentā¦ā
Some good reads to explore:
A profile of Dario Amodei suggests he thinks AI is so powerful that only he and those who share his beliefs should direct its development.
Why does America hate AI? Is it American individualism, wonders Gillian Tett.
On the productivity question
My view is that AI has yet to show up systematically in the aggregate productivity statistics ā nor would we expect it to yet. In the PC era, it took fourteen years for computers to make their mark on headline productivity numbers. But individual companies started to report gains several years earlier. AI will follow the same pattern, but might be faster.
Now, Erik Brynjolfsson argues that āweāre turning the corner on productivity,ā as US productivity grew at 3.6% in the past quarter.
Productivity data is quite noisy, so I donāt think we can attribute this to AI with much confidence yet.
But the evidence that individual firms intend to get results is increasing. A&O Sherman, a big law firm, is putting 3% of billable time into AI ā about $1.1bn over five years. The ambition is āredesigning the service completely.ā It echoes what I keep hearing in private discussions with bosses: small successes now, but a sense they know how to get large results over the next few years.
See also:


