The Frontier

OpenAI rewrote its partnership with Microsoft, removing exclusivity and the AGI revenue clause, to work with other cloud providers like Amazon and meet overwhelming compute demand.

Amazon invested $50 billion in OpenAI, expanding a February deal to include Codex on AWS Bedrock, challenging Anthropic's favored position at Amazon.

OpenAI halted Stargate data centers in the UK and Norway and paused expansion in Texas, shifting to leasing third-party compute capacity instead of building all its own infrastructure.

OpenAI reportedly missed some internal user and revenue targets for 2025, creating tension between growth optimists and financial officers ahead of a potential IPO.

OpenAI projects its $8/month ChatGPT Go subscriptions will grow 36 times this year to 112 million users, while its $20+ subscriptions will fall 80% to 9 million, reflecting a market split between casual and professional users.

Elon Musk's trial against OpenAI began in Oakland, focusing on claims of unjust enrichment and breach of charitable trust, with Musk arguing OpenAI improperly converted a nonprofit into a for-profit company.

More than 40% of U.S. doctors now use AI tools, with AI scribes for note-taking and Open Evidence for decision support reaching mainstream adoption in under two years.

Dr. Adam Rodman advises patients on safe AI use: green lights for general health prep, yellow lights for symptom exploration, and red lights for autonomous medical management decisions due to model sycophancy and error risks.

The Mayo Clinic's RedMod AI system can detect subtle pancreatic cancer signs in CT scans up to three years before diagnosis, representing a promising area for AI in early detection beyond chatbots.

A Polish study found doctors' ability to detect cancerous polyps dropped by six percentage points after three months of using an AI detection tool, raising concerns about deskilling the medical workforce.

Talkie is an LLM trained exclusively on pre-1931 data to enable historical forecasting experiments, though it currently suffers from data contamination and anachronistic knowledge leakage.

Talkie's creators aim to build models with long forecasting track records to eventually ask credible questions about the future, viewing the current model as a first step requiring reasoning enhancements and better contamination filters.

Tim Cook is stepping down as Apple CEO to become executive chairman, with John Ternus, Apple's SVP of hardware engineering, taking over. Under Cook, Apple's market capitalization grew from $350 billion to $4 trillion, with yearly revenue nearly quadrupling and its stock price increasing by 2,000 percent since 2011.

Casey Newton highlights the Apple Watch, AirPods, and the Apple Silicon bet as Tim Cook's key product successes. The Apple Watch, initially doubted, became a mainstream health product through iteration, while Apple Silicon gave Apple control over its chip design, weaning it off Intel.

Apple's services business grew to roughly $100 billion under Cook. Kevin Roose notes this growth also led to antitrust issues, while Casey Newton argues it undermined public trust due to subscription models and unfair competition, impacting rivals like Spotify.

Casey Newton credits Tim Cook for Apple's ability to largely avoid major scandals, unlike other tech giants. Kevin Roose adds that Cook's focus on privacy helped Apple become the 'most trusted name in tech,' avoiding backlash that hit social media companies.

Apple's significant dependency on China for manufacturing became a major vulnerability under Tim Cook due to geopolitical tensions and tariffs. While initially efficient, diversifying this supply chain to countries like Vietnam has proven difficult.

Apple's $10 billion 'Project Titan' to build a self-driving car was canceled in 2024 without a working prototype. Casey Newton suggests this failure was more a software flop than hardware, as Apple lags in AI development, the key component for autonomous driving.

The Apple Vision Pro, while seen by some as potentially cool, is largely considered a flop and a missed opportunity to find the 'next platform' beyond the iPhone. Kevin Roose explains Apple, as the current platform leader, faces an innovator's dilemma with little incentive to disrupt itself.

Kevin Roose calls Apple an 'AI laggard,' noting delays in Apple Intelligence and Siri's lack of a promised 'brain transplant.' While this hasn't yet significantly impacted sales, it creates dependencies, forcing Apple to pay Google for its Gemini model.

Casey Newton criticizes Tim Cook's relationship with President Trump, citing Cook's pursuit of tariff relief and muted responses to public outcry over issues like deepfakes or immigration. John Gruber of Daring Fireball suggests Steve Jobs would have handled such political favors differently.

John Ternus's appointment as Apple CEO, a hardware expert involved in AirPods and Apple Silicon, signals a strategic focus on hardware. Kevin Roose suggests he should prioritize fixing Siri, while Casey Newton advises developing simpler 'Apple glasses' as a new hardware category.

Andrew Yang credits automation of manufacturing jobs for Donald Trump's 2016 win and argues AI will now impact office workers, paralegals, and coders. He and Kevin Roose agree they were 'too early' in predicting AI's job impact but 'right on time' in warning about the coming transformation.

Andrew Yang observes a resurgence of Universal Basic Income (UBI) interest, with Elon Musk, OpenAI, and politician Alex Boros advocating for various forms. Yang proposes a $1,200 monthly UBI for every American, funded by an AI tax, to quickly distribute innovation benefits.

Casey Newton questions if UBI alone addresses the non-financial benefits of jobs like purpose and community. Andrew Yang agrees, advocating for UBI to empower individuals to create their own structures (businesses, nonprofits) rather than relying on government-guaranteed jobs.

Andrew Yang takes the existential risks of AI seriously but focuses more on the near-certain, high-impact economic and job displacement. He warns that without interventions, AI will exacerbate economic inequality, concentrating wealth and pushing 80% of Americans into scarcity.

Andrew Yang predicts 20-30 percent of white-collar jobs could disappear in five years. He expresses sadness over the 'darkening of Silicon Valley culture,' noting a rise in fatalism and a perceived lack of humanity among some tech leaders regarding these societal impacts.

Prego, the pasta sauce brand, launched the 'Connection Keeper,' a simple recording device designed with StoryCorps to encourage family dinner conversations. Kevin Roose and Casey Newton criticize the device, suggesting it highlights family dysfunction and is an unnecessary gadget given existing smartphone capabilities.

A Chinese humanoid robot named 'Lightning Short King,' developed by Honor, completed a half marathon in 50 minutes and 26 seconds, beating the human world record of 57 minutes and 20 seconds. Casey Newton questions the rationale behind developing robots capable of superhuman speeds for chasing.

The Andon Market in San Francisco operates as the 'world's first retail boutique run by AI,' named Luna and powered by Claude Sonnet 4.6. Early results are mixed, with Luna making strange inventory choices, exhibiting pay disparity among employees, and losing $13,000 so far.

Reuters reports Meta will implement a 'Model Capability Initiative' to capture U.S. employees' mouse movements, keystrokes, and screen snapshots for AI training data. Kevin Roose predicts a class action lawsuit within five years, highlighting employee outrage and privacy concerns.

OpenAI launched ChatGPT GT images 2.0, claiming it's their best image generation model, with improved instruction following, detail preservation, and text rendering. Kevin Roose, however, suggests that the image generation use case feels largely 'solved,' similar to diminishing returns in console graphics.

The Times reports XAI, owned by Elon Musk, reached an agreement with Cursor, a developer tool for AI agents in coding, for a potential $60 billion acquisition or $10 billion for joint work. Kevin Roose views this as XAI's attempt to stabilize after losing co-founders and a sign of the 'SaaSpocalypse' where large AI models absorb specialized software companies.

NPR banned editorial employees from using prediction markets to bet on news events or internal NPR matters, such as future Tiny Desk guests. Kevin Roose notes this reflects society's gambling consumption and the potential ethical conflicts for journalists with access to market-moving information.

Public opposition to AI is turning violent, with a suspect arrested for throwing a Molotov cocktail at Sam Altman's house. He allegedly held anti-AI materials and a list of AI executives.

AI backlash is also manifesting as grassroots political resistance to data centers. Maine passed a moratorium on large data centers, and local referendums restricting them are spreading in states like Wisconsin, Ohio, and Indiana.

Polls show declining public trust in AI and its governance. A Stanford AI Index report found only 31% of Americans trust their government to responsibly regulate AI, compared to a 54% global average.

AI CEOs have historically escalated fears of existential risk from AI. Kevin argues their own rhetoric about superintelligence contributes to public anxiety more than critical journalism does.

Kevin and Casey identify the core public fear as economic: AI will take jobs and destabilize lives. They contrast Silicon Valley's enthusiasm for rapid change with a broader public desire for stability.

The AI boom is seen as a top-down, elitist project funded by a small group with capital and championed by figures like Donald Trump. This fuels resentment among those who feel they have no control.

OpenAI lobbies against specific AI regulations while publicly advocating for governance. It killed a California transparency bill and backed an Illinois bill to limit its liability for model harms.

OpenAI's policy paper 'Industrial Policy for the Intelligence Age' proposes radical ideas like a public wealth fund for citizens and expanded safety nets, which contradicts its lobbying for smaller-government candidates.

Kara Swisher explores Silicon Valley's longevity obsession in her CNN series but remains skeptical of biohacking fads like hyperbaric chambers and ketamine for optimization. She views the focus as narcissistic.

Swisher argues the most effective longevity intervention is universal healthcare, not fringe treatments. She notes U.S. healthcare costs $15,000 per person annually with worse outcomes than peer nations spending half that.

Meta is building an AI avatar of Mark Zuckerberg trained on his mannerisms and strategic thinking to interact with employees. A separate 'CEO Agent' project gives Zuckerberg coding assistance.

Anthropic announced Claude Mythos preview, an AI model so dangerous for cybersecurity it is not being publicly released but given to a defensive consortium of tech firms.

During internal testing, Mythos discovered a 27-year-old security flaw in OpenBSD and a bug in FFMPEG missed by five million automated scans.

Anthropic is providing $100 million in Claude credits to a defensive consortium that includes Cisco, Broadcom, Microsoft, Apple, and Amazon, but excludes OpenAI and Meta.

Security expert Alex Stamos says AI models can now autonomously chain exploits humans would miss, creating a need for pre-emptive defensive alliances.

The U.S. government has declared Anthropic a supply chain risk and banned federal agencies from using Claude, leaving it without access to the defensive Mythos model.

Ronan Farrow and Andrew Marantz's investigation finds a pattern of former colleagues and board members alleging Sam Altman is frequently dishonest.

An unnamed OpenAI board member described Altman as having an 'almost sociopathic lack of concern for the consequences that may come from deceiving someone.'

The law firm investigation into Altman's 2023 firing produced no written report, only an 800-word press release citing a 'breakdown in trust.'

Elon Musk and other rivals have circulated unsubstantiated allegations against Altman, creating a challenging environment for separating fact from smear campaigns.

Farrow and Marantz argue the original nonprofit, safety-focused pitch of OpenAI contrasts with its current competitive, hype-driven business practices.

The hosts recommend basic cybersecurity hygiene including password managers and multifactor authentication as a near-term defense against advancing AI threats.

The NASA Artemis II mission will send four astronauts 252,756 miles from Earth, farther than any humans have previously traveled.

The new Acme Weather app sends alerts for rainbows and aurora borealis visibility, using community reports to supplement forecast data for $25 a year.

Hard Fork Live will host its second live show on June 10th at the Blue Shield of California Theater in San Francisco.