Untangling Web3

🎙 Untangling Web3 The podcast that simplifies tomorrow's technology, today. Follow and subscribe to join our journey to untangle Web3.

What makes this significant: Spain isn't just investigating Polymarket. Kalshi is regulated by the CFTC in the US. It cleared years of legal hurdles to operate. Spain's regulator is treating both platforms the same way - which means the blockchain isn't the issue. The bet is.

Spain blocked Polymarket and Kalshi this week - not permanently, but for 3-4 months while regulators decide if prediction markets count as gambling. A temporary block with a deadline is still a block. And the clock is running.

Every EU regulator is now watching Spain's next move. If Spain rules "gambling," it hands neighboring countries a ready-made template. 3-4 months from now, this stops being Spain's problem and starts being the entire European market's answer.

Should the state be innovating in product development, or is that the private sector's domain? The debate is heating up as AI and Web3 push for greater democratization in finance. Learn how these forces could empower individuals to be more autonomous and sovereign.

Prediction markets built their reputation on "the crowd knows best." That argument collapses the moment someone at the table has a classified briefing. Congress is now asking whether these platforms can police that - and neither Polymarket nor Kalshi has a clean answer yet.

Polymarket and Kalshi are being investigated by Congress for insider trading. The accusation: federal employees with security clearances may be betting on political outcomes they already know. There's no law stopping them. The STOCK Act covers stocks - not prediction markets.

Rep. James Comer is demanding internal trading records from both platforms. The ask is pointed: show us who's winning, and how. If a government official knows a policy decision before the public does, a well-timed bet pays out with zero legal exposure today.

Musk says SpaceX is "in discussions with other companies to do the same." The most powerful leverage point in AI right now isn't the models. It's the buildings with the electricity.

Anthropic is paying SpaceX $15 billion a year for computing power. Not Amazon. Not Google. SpaceX - the company whose main AI product, Grok, competes directly with Claude. Anthropic confirmed the number to Wired.

The deal runs through May 2029 at $1.25 billion a month, giving Anthropic access to Colossus and Colossus II in Tennessee and Mississippi. Anthropic's Q2 2026 revenue is expected to top $10 billion. It's still spending 37% of that - annually - just to keep its models running.

Running a tech event? Untangling can help with: • live podcast recordings • panel hosting + moderation • guest speaking • content capture for clips + social We bring hands-on experience making complex tech topics into compelling live conversations around AI, blockchain, digital identity, IoT, and more. Looking for support at your next event? Get in touch: theuntanglingweb3podcast@gmail.com

SpaceX's IPO filing made that number public for the first time. $15 billion annually - just for Anthropic to access data centers. That's not a vendor relationship. That's a dependency. And SpaceX's owner is also building a competing AI through xAI.

Anthropic is spending $15 billion a year on compute. That's more than most Fortune 500 companies earn in profit - and it's going to SpaceX. Not AWS. Not Google. A rocket company is now one of the biggest AI infrastructure landlords on Earth.

Anthropic is paying its rival's parent company $15B a year to keep the lights on. If SpaceX raises prices, changes terms, or prioritizes xAI - what exactly can Anthropic do? That's not a hypothetical risk. It's already the current situation.

Last week on Untangling Web3, we sit down with Pierre Person (@Pierr_Person) CEO & Co-Founder of Usual Labs (@usualmoney) and Fira Money (@Fira_Lend), and former French MP - to unpack how finance itself is being rebuilt around programmability, automation, and onchain infrastructure. Check out the full episode now: Youtube: youtu.be/lW_9duAiVEQ?si… Spotify: open.spotify.com/episode/144cEa… Apple Podcast: podcasts.apple.com/gb/podcast/unt… @AlecJBurns @JackODavies_

Personal finance is about to get a major upgrade. Imagine having an AI agent with deep market knowledge, leveling the playing field between retail investors and institutional giants. This could democratize finance, giving everyone the capacity for more informed decisions.

The stakes behind the ruling: youth unemployment in China sits at 17%. Over 200 million workers are already stuck in low-wage gig roles. Courts are being used as a pressure valve before that number climbs higher. Matt Sheehan at Carnegie put it plainly - Beijing wants AI everywhere and wants zero job losses. Those two things cannot both be true.

China just ruled that firing workers to hire AI counts as optional cost-cutting - not business necessity. Three separate courts have now sided with displaced workers. Employers cannot use automation as legal grounds for mass layoffs. That one distinction could reshape how every company in China prices its AI roadmap.

So now automation has a new line item: legal liability per dismissed worker. Companies will not stop building. They will just absorb the lawsuits as a cost of doing it. That is not job protection. That is a delay with paperwork.

Every org with a "we exist for good" charter just got a reality check. If OpenAI's shift from nonprofit to for-profit giant doesn't count as a breach - what does? Courts won't govern AI ethics. So who will?

Elon Musk sued OpenAI for $150 billion - and lost. He co-founded the company, left the board in 2018, then watched it take billions from Microsoft and become the most dominant AI product on earth. A jury saw all of that and said: no case.