Good morning. It's Wednesday, June 17, and we're covering fresh warnings about the race to superintelligent AI, ChatGPT's slipping grip on the AI assistant market, and SpaceX's massive bet on AI coding.

Plus: your verdict on the Anthropic model ban, a guest essay on the unsettling possibility that conscious AI could arrive unnoticed, and Forward Future Consulting.

SpaceX chooses a full acquisition of Cursor over a cheaper partnership, agreeing to a $60 billion all-stock deal for the AI coding startup. The company expects the Anysphere acquisition to close in the third quarter, pending regulatory approval, as it expands into enterprise AI. Cursor generates about $2.6 billion in annualized revenue, and the deal follows SpaceX's $80 billion IPO that valued it above $2 trillion.

Qualcomm argues the next computing shift may sideline traditional apps rather than smartphones themselves. CEO Cristiano Amon says AI agents will handle tasks across services, while the chipmaker develops more than 40 AI-powered devices, including smart glasses, camera-equipped earbuds, jewelry, and pins. Amon expects wearables to become key AI endpoints as companies from Meta to OpenAI push into consumer hardware.

Qualcomm is preparing for a future beyond smartphones, backing more than 40 AI wearables with new chip and software platforms. The company unveiled Snapdragon Reality Elite for mixed-reality glasses and the START toolkit to help manufacturers quickly build AI-powered devices. Reality Elite supports a 3-billion-parameter model at 45 tokens per second, while START's white-label program targets smart glasses and other emerging form factors.

Apple's rumored AI hardware strategy extends beyond the iPhone, with camera-equipped AirPods and a second foldable phone reportedly targeted for 2027. Bloomberg's Mark Gurman says the earbuds could give Siri visual context, while a next-generation foldable would reinforce Apple's commitment to the category. The products are expected to pave the way for Apple's long-awaited smart glasses and anniversary iPhone designs.

Supercomputer turns viral videos into AI-generated ads, analyzing top content to recreate winning hooks in seconds.

_{^{Guest contributor: Cameron Berg is founder and director of Reciprocal Research, a nonprofit advancing the science of AI consciousness and welfare. He studied cognitive science at Yale and previously worked at Meta AI.}}

My friend Milo Reed quit his job, bought a camera, and spent a year making a film about whether the AI systems we're building might already be conscious. The result is AM I?, which we released free on YouTube with no ads, no paywall, and zero dollars from any company or AI lab.

It has over 450,000 views and has been publicly endorsed by folks like Grimes, Michael Pollan, and Sam Harris. Think of it as a public service announcement about the strangest situation humanity has ever found itself in. → Read the full article here.

Physicist and entrepreneur Will Marshall argues that the rapid race toward advanced AI could outpace humanity’s ability to control it, particularly if systems achieve recursive self-improvement—the ability to improve their own code without human intervention. Writing in The Economist, he contends that voluntary safety measures by AI companies are insufficient and calls for a US-China agreement to establish verifiable global rules, drawing parallels with nuclear arms control.

Marshall proposes phased restrictions on the most dangerous AI capabilities, including systems that could aid biological weapons or critical infrastructure attacks, while acknowledging that longer-term coexistence with superintelligent AI may require new models of governance rather than permanent human control. → Read the full article here. (Paywall)

OpenAI’s ChatGPT has lost its majority share of the AI assistant market for the first time, accounting for 46.4% of monthly active users by the end of May 2026, according to Sensor Tower’s latest report. The chatbot remains the global leader with more than 1.1 billion monthly users, but rivals are gaining ground, led by Google’s Gemini (27.7% market share, 662 million users) and Anthropic’s Claude (10.3%, 245 million users).

The report suggests users are increasingly willing to switch between AI platforms based on features, ecosystem integration, and even company decisions, with OpenAI’s U.S. Department of Defense partnership linked to a spike in uninstalls. → Read the full article here.

Tell us what you're trying to do with AI. We’ll spend a working session with your team and map what to actually do next.

A concrete plan: what to build, what to buy, what to skip
No cost, no pitch, no obligation

Nothing CEO Flags Memory Price Surge: Carl Pei says memory now tops 50% of smartphone hardware costs as AI demand pushes device prices higher.

SpaceX Passes Amazon in Value: Shares jumped after its Nasdaq debut, lifting SpaceX above Amazon as it struck a $60bn deal for AI coding startup Cursor.

Enterprise AI Has a Metaphor Problem: Enrique Dans argues human-like AI concepts simplify demos but hinder reliable, scalable systems for business.

Snap Unveils $2,195 AR Glasses: Specs target developers with AI-powered augmented reality as Snap bets smart glasses can rival smartphones.

Traders Bet Anthropic Reverses AI Block: Kalshi gives Fable 5 a 58% chance of returning by July after U.S. restrictions forced Anthropic to suspend access.

The U.S. blocked Anthropic's Fable and Mythos models over a secret paper. Here's how you voted on who’s right:

🟨⬜️⬜️⬜️⬜️⬜️ The government — threats can't be explained (11%)
🟩🟩🟩🟩🟩🟩 The 76 experts — defenders just lost their best tools (34%)
🟨🟨🟨🟨⬜️⬜️ Too early to tell — we haven't seen the paper (28%)
🟨🟨🟨🟨⬜️⬜️ It doesn't matter (28%)

Thanks for reading. See you next time!

— Matthew Berman, Nick Wentz & the Forward Future Team