Dom Hallis
4 Aug 2026Dom Hallas: What's Really Happening in US-China AI Competition | FULL INTERVIEW
Welcome, Dom Hallis — AI Politics Temperature Check: US, UK, Europe
The hosts welcome Dom Hallis to their new in-person studio and note the announcement that Double O House is back. Dom gives a temperature check on AI politics: the industry letter responding to Chinese open-source models (Jensen Huang's first tweet, Anthropic conspicuously absent), a same-day White House meeting on pre-release testing, and the shift away from David Sacks and Sriram toward a more regulatory posture under Bessent, Wiles and Lutnick. He argues the turn is downstream of China geopolitics and the security-versus-growth policy split, points to the UK's 'monitoring the situation' approach via the AI Security Institute versus Europe pushing ahead with rules, and flags rising US public and cross-party pushback on data center construction, including DeSantis in Florida.
The UK–US Adoption Gap & Mistral's Forced Adoption
On adoption, the hosts note Kanishka's move to the Cabinet Office with the PM wanting him at the weekly table, and Dom contrasts the US 'do a thing, fix the problems later' permissionlessness with slow-adopting FTSE 100 and European corporates, which is why European startups often build product at home but sell in the US. He illustrates with English industrial history: Manchester's brutal-but-transformative growth that Engels wrote about, and the Bradford Wool Exchange where Ruskin refused to bless a temple to excess. He then offers Mistral as the counterexample of French industrial policy effectively force-feeding its products to CAC 40 companies — good for the business, but risking 'bad customers make bad businesses' and weaker frontier capability.
AI Geopolitics & the New Tech Diplomacy
After the hosts mention their conversations with Sam Hogg on the UK–China AI relationship, Dom describes the emerging tech-diplomacy role: policy chiefs like Kent Walker and Nick Clegg carry 'global affairs' titles, governments now post tech envoys to Silicon Valley, and companies' R&D budgets rival nation states. He recommends the book 'Apple in China' for how firms navigate Beijing's intentionally vague rules, which are written to be broken so leverage can be extracted. He closes with an anecdote from a tech conference in Egypt, where an audience member asked what to do when a minister asks for a bribe, drawing on his earlier stints in Brussels and India.
Should Governments Nationalize Frontier Labs?
On whether governments should nationalize frontier labs, Dom contrasts the post-Cold War economy with the Attlee government's post-1945 nationalizations — down to the entire road haulage industry — but argues frontier AI requires an ecosystem that the state cannot monopolize the way it traditionally monopolized violence. He walks through the National Security and Investment Act and its US equivalent, which lets the state block hostile-state acquisitions but creates dual-use headaches like sonar-based fish-finding AI and facial recognition for chicken welfare. A founder's knife analogy ('cut fruit or kill someone') and centuries of debates over restricting expression frame where lines might be drawn.
Open Weights vs Closed & Selling the Growth Story
The hosts weigh excitement over OpenAI's math breakthroughs — and hoped-for spillovers into biology and chemistry, maybe even a cancer cure — against open-weight pacing concerns, and recall a recent founder guest arguing open weights will eventually own most of the market, before a brief Polymarket plug. Dom says the market outcome is unpredictable and the open/closed fight will increasingly be a negotiation in which politics and public consent matter. He argues the industry's 'up and to the right' growth story is no longer sufficient — pointing to Andy Burnham's 'good growth in every postcode,' the social movements and model towns like Saltaire born of industrial Manchester's upheavals, and the risk of 'shaving the edges off success' and ending up with a $7B company instead of a $10B one without fixing public sentiment.
Railway-Boom Parallels, Slop Backlash & Growing the Pie
Asked whether AI resembles the railway or industrial revolution, Dom cites the recent New York Times chart showing AI capital investment rivaling past booms like the railways. He notes industrial benefits took time to be felt, so a 'slop' backlash was predictable — recalling how OpenAI quickly killed its slop-video app — and says the sector must show real value rather than try to stop the train. He ends on growing the pie before fighting over distribution: Microsoft pays roughly $20M a year in property taxes to one small Wisconsin town, and he argues the UK needs fiscal devolution of the kind Burnham is exploring so communities actually see the upside of data centers.
Outro & Thanks
The hosts wrap up a minute over time, thank Dom for coming on, and welcome him back any time before signing off by introducing Dom Hallis to the audience.
19 Jan 2026Europe’s Newest Unicorn, Hiro Capital Deep Dive, Thinking Machines Drama, Listen Labs does $100M
Dom Hallis on Nick Clegg, Hiro Capital & Europe's Growth-Cap Gap
Dom Hallis, executive director of the Startup Coalition, joins from London to unpack Nick Clegg joining Hiro Capital and Yan LeCun's growing European presence, including his advisory role at Prior Labs. He argues Europe should build on its own strengths—fintech, research universities, adaptability to fragmented markets—rather than clone Silicon Valley, and breaks down the growth-stage "valley of death," capital supply, pension fund reform, and why different industries need different return profiles, tangenting into his Sharia-compliant pension fund outperforming everything else.