DX Today | No-Hype Podcast & News About AI & DX

Send us Fan Mail SpaceX Becomes a Compute Landlord: The $6.3 Billion Reflection AI Deal and the New Economics of Open Source Compute - June 24, 2026 Chris and Laura dig into SpaceX's reported computing power deal with open source startup Reflection AI, worth up to $6.3 billion through 2029. They unpack why a rocket company is now renting Nvidia GB300 capacity at the Colossus 2 data center, why a $25 billion startup is committing $150 million a month before shipping a public flagship model, a...

Send us Fan Mail Project Glasswing: How AI Now Finds Vulnerabilities Faster Than the World Can Patch Them - June 23, 2026 Anthropic's Project Glasswing used an unreleased frontier model, Claude Mythos Preview, to uncover more than 10,000 high and critical severity software vulnerabilities in a single month. In this deep dive, Chris and Laura explore the most surprising twist of all: the bottleneck in cybersecurity has inverted from finding flaws to verifying, disclosing, and patching them fa...

Send us Fan Mail AI Agents Are Becoming Employees: NewCore's $66M Bet and the Machine Identity Crisis - June 22, 2026 Chris and Laura dig into NewCore's $66 million stealth launch at a $300 million valuation and the fast growing problem it targets: as enterprises deploy autonomous AI agents, those agents now outnumber human staff by dozens to one and live in an identity gray area that legacy security tools were never built to govern. They unpack the split key architecture, the Agentic Skill ...

Send us Fan Mail FERC Orders Six Grid Operators to Fast Track AI Data Center Power While Protecting Ratepayers On June 18, 2026, the Federal Energy Regulatory Commission issued six show cause orders directing PJM, MISO, SPP, CAISO, ISO New England, and NYISO to accelerate grid interconnection for AI data centers while barring cost shifting onto ordinary ratepayers. Rick Spair and Laura unpack the 30 and 60 day deadlines, the Texas carve out, the Maryland ratepayer fight, and why the race for...

Send us Fan Mail Google's Always-On Search Agents: How Information Agents Turn Search Into a Service That Works While You Sleep At Google I/O 2026, Google unveiled "information agents" in AI Mode, persistent background processes that monitor the entire web around the clock and send synthesized push updates without you ever opening the search box. Chris and Laura dig into how this flips search from pull to push, why it launched gated behind the AI Ultra tier, and what always-on monitoring mea...

Send us Fan Mail DX Today AI Daily Brief - Friday, June 19, 2026 Today's briefing: OpenAI loads up on talent, hiring transformer co-inventor Noam Shazeer and former White House AI policy official Dean Ball ahead of its IPO. Amazon moves to sell its Trainium chips, challenging Nvidia. Anthropic restricts its newest Claude models for foreign nationals under US export controls, igniting debate in India. xAI's Grok lands natively on Databricks Agent Bricks. Elastic acquires DeductiveAI for up to...

Send us Fan Mail AI's Two-Track Labor Market: Inside the PwC 2026 Global AI Jobs Barometer and the 62% Skills Premium - June 18, 2026 PwC's 2026 Global AI Jobs Barometer, drawn from roughly a billion job ads, finds AI is splitting work into two tracks: jobs "professionalised" by AI grow twice as fast with 42% faster wage growth, while the AI skills premium climbs to 62%. Chris and Laura unpack what it means for workers, entry-level careers, and the companies pulling ahead. Hosted by Chris a...

Send us Fan Mail The AI Infrastructure Capacity Crisis: Microsoft Borrows AWS to Keep GitHub Online and Even Cloud Giants Run Out of Compute - June 17, 2026 Chris and Laura dig into the week's quietest but most consequential AI story: the physical limits of the buildout. Microsoft confirmed it is routing GitHub traffic through rival Amazon Web Services after AI coding agents pushed availability to roughly 88.4 percent, while Google agreed to pay SpaceX 920 million dollars a month for compute...

Send us Fan Mail The First AI Formulated Drug Reaches the Clinic: Inside Quotient Sciences' Phase I Milestone and Pharma's 2026 Inflection Quotient Sciences, working with Intrepid Labs' machine learning, has begun what is reported as the first human clinical study of an AI designed drug formulation, after MHRA approval in the UK. Chris and Laura unpack the crucial difference between an AI designed molecule and an AI designed formulation, and what it means that there are now 200+ AI originate...

Send us Fan Mail DX Today AI Daily Brief - Monday, June 15, 2026 Anthropic takes its most advanced Fable 5 and Mythos 5 models offline worldwide after a US export-control order, while Moonshot AI counters with an open-weights release of Kimi K2.7 Code. Mistral closes in on a roughly 20 billion euro valuation, Jeff Bezos returns to the founder's seat with a 12 billion dollar AI engineering startup called Prometheus, and Google sues a China-based cybercrime group for abusing Gemini. Visa wires...

Send us Fan Mail TurboQuant: Google's 6x KV Cache Compression and the Quiet Economics of Long Context AI Google Research's TurboQuant compresses the LLM key value cache to roughly three bits per coordinate with near zero accuracy loss, delivering at least six times less memory and up to eight times faster attention on NVIDIA H100 GPUs. We unpack how its two stage design pairs a training free random rotation with a one bit correction step, why a 70B model's 128K context cache shrinks from abo...