The Morning Download: Confronting AI's Economic Impact

Good morning. Maine is likely to become the first state to freeze the construction of large new data centers, a sign of a growing community backlash across the U.S.

Venture capital pioneer Vinod Khosla, an early investor in OpenAI, told me during a recent conversation that such public concerns must be taken very seriously. The technology is evolving so fast that massive job dislocations are on the horizon and a political backlash could derail the realization of AI’s promise in a range of fields.

Lawmakers in more than 10 states have proposed temporary bans on data-center construction this year, and dozens of county and city governments have already passed such measures, the WSJ reports. Take a look at its map of bans that have been proposed or passed.

Khosla told me that AI already outperforms highly trained human professionals in a range of tasks across fields from medical diagnosis and materials science, such as the design of new and better magnets. The pace of AI innovation is accelerating, and “the last two years of progress, we’ll see in one year,” he said.

Adoption will be slower, though.

“It will take a long time for the technology to diffuse. That’s a normal process, and so I expect what a few people know, what 1% of the population knows, will take until 2030 for 90% of the population to understand,” Khosla said.

Looking ahead, the most significant barriers to adoption won’t be in the areas of innovation or even security. They will be social, according to Khosla, who has proposed scrapping income taxes on people earning less than $100,000 a year, and facilitating those changes by taxing capital gains at the same level as income. He raised those ideas last month at the Hill & Valley Forum.

Fundamental modernization of the social contract is necessary as AI evolves during the next few years, according to Khosla.

“Mark my words, in ‘28 it will be a huge issue and in 2032 it may be the only issue in a presidential election. I think the political problem is far bigger than most people realize in the adoption of AI,”  he said. “There'll be lots of excuses, but the biggest reason will be jobs. That's why I made some very radical proposals .. I do think we have to take a very radical look at taking care of people before we start impacting their lives.”

How concerned are you that AI will displace your job in the next few years? Let us know. 

The Journal reports that Anthropic is in talks to invest $200 million in a new venture with General Atlantic, Blackstone and Hellman & Friedman. The new company would serve as an Anthropic consulting arm teaching businesses how to incorporate its AI tools into operations.

OpenAI is also in talks to form a rival joint venture with private-equity firms.

The appeal of PE. Companies backed by private-equity firms are appealing customers to AI firms in part because their owners are already trying to cut costs, the Journal explains. PE firms can also push technology decisions across their entire portfolios of investments.

While investment firms look to automate more and more business processes with AI, this guy wants to automate a whole industry. 

Pompliano tells the WSJ that ProCap Insights was built over the course of two weeks and cost only a couple thousand dollars. One employee oversaw the project. “My guess is that the entire industry is moving this way,” he said.

Meanwhile ... more older workers opt to retire rather than learn AI

The share of Americans over 55 years old in the workforce has slipped to 37.2%, the lowest level in more than 20 years. Rising home equity and stock-market returns is driving some of the decline, economists and retirement advisers say. Recent retirees tell the WSJ that AI is another factor. Having worked through the PC revolution, the dawn of the internet and smartphones, they are done. 

Broadcom will develop and supply custom AI chips for Google and additional computing capacity to Anthropic in an expansion of the strategic collaboration between the three companies, WSJ reports.

More highlights:

• Broadcom said it will supply Google with custom Tensor Processing Units, along with networking and other components for Google’s next-generation AI data center racks through up to 2031 as part of a supply assurance agreement.
• Anthropic will access about 3.5 gigawatts of TPU-based computing capacity beginning in 2027 as part of its commitment for multiple gigawatts of compute capacity, Broadcom said.

Bezos startup hires xAI co-founder. A secretive Jeff Bezos owned project code-named Project Prometheus is on a worldwide hiring spree, scooping up AI researchers and infrastructure experts to work on AI systems that understand the physical world, the FT reports. Recently, OpenAI's Kyle Kosic, an xAI co-founder, joined the effort.

Meta to open source versions of its next AI models. Meta is preparing to release the first new AI models developed under Alexandr Wang, with plans to eventually offer versions of those models via an open source license, Axios reports.

OpenAI, Anthropic, Google unite to combat model copying in China. The firms are sharing information through the Frontier Model Forum, an industry nonprofit that the three tech companies founded with Microsoft in 2023, to detect so-called adversarial distillation attempts that violate their terms of service, Bloomberg reports.

Samsung forecasts record first-quarter operating profit. The company forecast more than an eightfold jump in first-quarter operating profit from a year earlier, pointing to record earnings driven by its flagship semiconductor business on robust chip demand fueled by AI, WSJ reports.

Amazon and U.S. Postal Service reach delivery deal. Instead of reducing the number of packages Amazon ships through the Postal Service by two-thirds by this fall, as outlined in an earlier proposal, the two sides now have a tentative deal that will result in a 20% reduction, WSJ reports.

Iran rejected a proposal from the U.S. and regional mediators to end hostilities for 45 days in exchange for opening the Strait of Hormuz. (WSJ)

The Artemis astronauts have gone the distance—the longest in human history, in fact. Just before 2 p.m. ET Monday, the astronauts on NASA’s Artemis II lunar mission rode their Orion craft into the record books, flying more than 248,655 miles from Earth. (WSJ)

The Supreme Court on Monday paved the way for a lower court to erase the criminal co