Harvard AI Diagnoses, PsiQuantum’s $500M Quantum Race, Mistral vs OpenAI Geopolitics, NYC AI Regulation Clash

**Adam Nagus** (0:05)
Welcome to Digimasters Shorts. We are your hosts, Adam Nagus.

**Carly Wilson** (0:09)
And Carly Wilson, delivering the latest scoop from the digital realm.
A new study from researchers at Harvard and Stanford found that ChatGPT outperformed hundreds of physicians in a complex diagnostic challenge using written case histories and real patient data.
The results published in Science unsettled many doctors, even as lead author Adam Rodman cautioned that the experiment does not prove AI is ready for routine medical practice. Despite those warnings, hospitals across the country are rapidly adopting generative AI tools, many of which have not been approved by the FDA for clinical use. Physicians say the rollout feels unprecedented in a field that typically moves slowly to protect patient safety. While some research suggests AI can help diagnose rare diseases, other studies show that incorrect AI output can mislead doctors and fail to warn patients about medical emergencies. Errors are also appearing in administrative tasks, with some hospitals using AI to draft patient messages that physicians describe as inaccurate or absurd. Many of these tools avoid FDA oversight by classifying themselves as clinical decision support or wellness apps, rather than medical devices. Tech companies insist their chatbots do not provide diagnoses or treatment recommendations, even as they market features that interpret lab results and medical scans. Some experts argue that advanced medical chatbots should be licensed and supervised more like physicians, though federal approval efforts could take years. With 80% of doctors reportedly using AI and patients increasingly following suit, the technology is embedding itself in health care faster than regulators can respond.

**Adam Nagus** (1:42)
Illinois Governor JB. Pritzker and PsiQuantum CEO Victor Peng are betting big on a quantum computing hub that could redefine America's next tech boom.
On the south side of Chicago, a $500 million state-backed innovation campus is rising ARAG-E, anchored by a 300,000-square-foot facility that aims to house the world's first utility-scale fault-tolerant quantum computer. PsiQuantum, valued at $7 billion and backed by major federal funding, is racing to build a machine capable of solving problems beyond the reach of today's supercomputers. Unlike competitors taking incremental steps, the company is pursuing a full-scale fault-tolerant system from the start, using photonic chips manufactured with semiconductor giant Global Foundries. The stakes are both technological and political, as Pritzker positions Illinois to compete with Silicon Valley and China in the global quantum race. Determined not to repeat the talent exodus of the 1990s, the governor is building an ecosystem designed to keep quantum startups and graduates rooted in the state. The project has already attracted major players including IBM and DARPA, while promising new jobs ranging from quantum scientists to cryogenic engineers.
Though timelines have stretched and technical hurdles remain, Peng calls the first machine only the end of the beginning of a broader transformation.
Even before completion, the partnership between PsiQuantum and Illinois has catalyzed a fast growing quantum corridor with ambitions as big as the technology itself.
At a G7 lunch in Evian this week, Mistral CEO Arthur Mensch sat alongside leaders of OpenAI, Google DeepMind and Anthropic just days after the US government ordered Anthropic to cut foreign nationals off from its most advanced models. The shutdown turned a long running European warning about AI dependence into a live geopolitical lesson. Online, that anxiety morphed into satire with Le Chaton Fat, a fictional Mistral supermodel boasting 1,000 meows per second and fake charts crushing its American rivals. The joke resonated because it captured a serious point. A European AI that cannot be switched off by Washington suddenly feels less like a slogan and more like a safeguard. Mench has argued for years that Europe must control its own AI infrastructure, recently warning lawmakers they have two years to avoid permanent dependence. France and the European Commission appear to be listening, backing Mistral for public sector use and framing the episode as proof of the need for technological sovereignty. Investors are responding too, with Mistral reportedly seeking billions in new funding at a sharply higher valuation. Yet sovereignty does not equal superiority, and Mistral still trails open AI and Anthropic in raw capability, scale and safety performance, including weak results in filtering Russian disinformation. The Anthropic episode handed Mistral its strongest political argument yet, but whether customers choose it for quality rather than nationality remains an open question.

**Carly Wilson** (4:36)
New York Assemblyman Alex Bors' run for Congress has become a multi-million dollar proxy battle within the technology industry over how to regulate artificial intelligence. A political group backed by investors tied to open AI has spent more than $7 million attacking Bors in the June 23 Democratic primary in Manhattan's 12th Congressional District. The group, Leading the Future, argues that AI regulation should be led by Congress and portrays Bors as aligned with a rival tech faction. Bors drew industry backlash after spearheading New York's Sweeping Raise Act, requiring major AI companies to report safeguards against catastrophic risks. In response to the attacks, political groups partly funded by Anthropic, maker of the Chatbot Claude, have poured more than $10 million into supporting Bors, with crypto billionaire Chris Larson pledging an additional $3.5 million.