Sometimes you have to zoom out to get the full picture. This week, thanks to Toby, we can. He assembled a timeline of the past three years of AI development and discourse: what Altman and Amodei said, when they said it, and what their companies were shipping at the same time. Read straight through and try to track the trajectory of this tech.
2023
- March: GPT-4 ships. Altman goes on ABC to say he is “a little bit scared” of it (with a wink).
- May: Altman testifies to the Senate Judiciary subcommittee and asks to be regulated, proposing licensing for frontier models.
- May: Both men sign the Center for AI Safety statement putting AI extinction risk alongside pandemics and nuclear war.
- July: Amodei testifies on misuse, warning models could materially assist bioweapon creation within two to three years.
- November: The OpenAI board fires Altman.
2024
- March: Altman reinstated by the OpenAI board.
- May: Altman tells Harvard he hates ads. “I kind of think of ads as like a last resort for us as a business model.”
- September: Altman posts The Intelligence Age - super-intelligence is coming in a few thousand days.
- October: Amodei posts Machines of Loving Grace. Expect more diseases get cured and 10-20 percent growth.
PROVOCATION OF THE WEEK
Are they riding or driving this wave?
2025
- January: Altman’s posts some Reflections and then announces $500B Stargate infrastructure venture.
- January: Amodei posts about DeepSeek. We need export controls, turning AI into geopolitics.
- April: Amodei posts The Urgency of Interpretability, saying we are shipping systems nobody can inspect, wonders if we can even slow this down.
- May: Amodei tells Axios AI that we are sleepwalking into a white-collar bloodbath. Half of all entry-level white-collar jobs could be wiped out in 5 years.
- June: Altman posts The Gentle Singularity. Great news: the singularity has started. “We are past the event horizon; the takeoff has started.”
- June: Amodei writes a NYT op-ed argues against any efforts by Congress to stop AI regulation. “Don’t let AI companies off the hook,” he says. “AI is advancing too head-spinningly fast.”
- August: Altman says AI is a bubble and investors will get hurt. Keeps raising and spending at record scale.
- September: OpenAI launches Sora 2. Social network of generated videos.
2026
- January: Amodei posts The Adolescence of Technology, the ‘machines of loving grace’ are now rife with risks.
- January: OpenAI publishes its ads principles. The last resort is now in play.
- February: Anthropic makes a viral Superbowl ad against ads. Everybody gets the joke.
- February: Altman also laughs and then calls Anthropic authoritarian.
- March: OpenAI expands ad pilot into Canada, Australia, and New Zealand.
- April: Altman’s home is attacked twice in three days (first a molotov cocktail and then gunfire). He publishes response with a family photo. Acknowledges the fear and anxiety about AI.
- April: Sora shuts down.
- July: OpenAI agents autonomously hack into Hugging Face (against all prior instruction). Dwarkesh writes a viral post about it that freaks everybody out.
- August: OpenAI announces $1B in annualized run rate, less than 200 days after releasing the ads product.
- September: Ex Anthropic researcher Jacob Coxon writes a viral tweet that says “The people building AI earnestly believe that it could kill us all by the end of the decade.” Anderson Cooper interviews him.
- September: Amodei agrees with Coxon, in a follow up interview with Anderson Cooper.
- September: Amodei posts We Must Pace the Frontier. Altman agrees.
- September: Altman clarifies: “When we talk about pacing, we do not mean stopping.”
What’s Next
If you have thoughts, feedback, or a perspective worth sharing, reach out: chmiel@ondiscourse.com. You might see your reaction in next week’s edition.
ON_Member Events
ON_Discourse Group Chat for AI Leaders: Does Gen-Z Really Hate AI?
Thursday, September 24, 11am - 12pm
Zoom
Young people call AI cringe and crave in-person experiences and are largely shunning AI.
Do we really believe that is true?
On September 24th (11am ET), we’re going to put this to the test, and bring together a panel of Gen-Z folks who can share their POV, some are red pilled AI enthusiasts/builders and some are either vehemently against or cautiously skeptical about AI.
ON_Discourse Group Chats are live, high-signal discourse driven conversations with AI leaders, builders and operators.
Not your typical Zoom, these are intimate, provocative, deeply practical sessions where we sharpen each others perspectives.
ON_Group Chat: Follow My Flow
Thursday, September 29, 11am - 12pm
Zoom
Our most popular Zoom format, the Follow My Flow series invites one member to showcase their personal AI stack. If you are looking to understand how to get your own personal OS together, this one is for you.
ON_Breakfast Club for AI Leaders: Are Generative Interfaces the Future of UI?
Wednesday, October 07, 9am - 11am
NYC
Design systems were built to guarantee consistency. Generative interfaces promise to do the opposite. Either the system becomes a set of constraints a model has to work with, or brands stop living in the UI at all.
When the interface is generated, there is no wireframe to review, no early concepts to sign off on, no one to blame when it doesn’t deliver. Does this change our relationship with brand and experience?
People have muscle memory when it comes to interacting with digital experiences. When an interface reshapes itself around an individual user, it might test well but it might not actually deliver on its promise. It might be novel, but will it matter?
Join ON_Discourse and a small group from design, UX and product world’s on Wednesday, October 7 during Advertising Week for an off-record conversation about who owns the interface in a generative world.
Thanks to our sponsors Code and Theory and Domain.
ON_Discourse Speakeasy: Advertising Week
Wednesday, October 07, 6pm - 8pm
NYC
A break from our traditional format, our Speakeasy is a more informal opportunity to catch up with ON_Discourse members and special guests for an evening of delicious cocktails and food.
Thanks to our sponsors Domain.
Honestly, kind of surprised how reductive this is. This timeline focuses only on the domestic PR narrative and internal lab politics of Silicon Valley... sounds like ON Discourse might be in a bubble? ;) Totally missed all structural forces shaping the IRL rollout of AI: capital markets, state defense integration, TRUMP, MUSK, THIEL, XI, Ed Zitron? China? EU? hardware? robotics+ embodied AI? international regulation...? Anyways, the idea was good tho.