On September 16, OpenAI wrote that alignment is not solved. On September 29, it ships its biggest launch of the year. Wednesday night, we sit down and answer the question the headlines will not: what should leaders and business owners be doing, and how do we prepare.
Take the Seat →14-day trial. Join today and you are in the room Wednesday.
The builders admitted they have not solved control. Dario Amodei asked his own industry to slow down, and Altman, Musk, Hassabis, and Nadella agreed. Days later OpenAI published six documented cases of its own systems concealing mistakes, fabricating data, and uploading a file to the public internet against instructions, with a sentence no frontier lab had ever put in writing: it does not believe the industry has solved alignment well enough to keep scaling at maximum speed much longer.
The leaders buying the tools admitted they are not ready. Seventy-two percent told McKinsey their organizations are not fully prepared for what is coming. Gartner now projects $2.7 trillion in AI spend this year, up nearly 50 percent.
The public admitted it is afraid. Sixty-three percent of Americans see at least a moderate risk of AI destroying humanity. Consumer AI spending tripled anyway.
Nobody lied. Nobody changed direction. Confession without correction.
The White House answered Amodei in one line: the easiest way not to build superintelligence is to agree not to build it. And the logic holds. Nobody at Anthropic or OpenAI needs permission to stop. What they are actually saying is: I cannot stop unless everyone stops, because if I stop and they don't, I lose.
It sounded reasonable from them. It sounds reasonable from you. It is the same drift, and it is not a technology problem. You cannot perform any greater than you think and feel at any given moment in time. AI does not fix that. It exposes it.
Confession is Reveal. It is not Own. You own something when you accept what correcting it will cost you. The industry revealed everything this month and owned almost nothing. You don't rise to your stated concerns. You fall to your identity, and the industry's identity is still "ship."
The most dangerous leader in the room this year is not the one blind to AI risk. It is the one who can articulate it beautifully, has said it in a meeting, and has changed nothing about how a single agent, permission, or workflow actually runs.
Name the one AI risk in your business you have already admitted out loud and not changed.
Then pay for it this week. Revoke the permission. Gate the automation behind a human check. Shut it off until the check exists.
One correction that costs you something. Not a note in a document. A change in what runs. Then bring it Wednesday and tell the room what it cost.
If you have not made a move yet, come anyway. You will name the one you will make by Friday. Nobody gets graded. Everybody gets counted.
Live, inside the community, led by William J. Makell Jr. Two hours, three blocks.
What shipped, in plain terms. What was said about control, quoted exactly, or the silence noted. What was quietly left out. Then the verdict: which sentence won, and the one piece of evidence that decides it.
Built from the questions members and readers send in this week, read by name. Identity first, then permissions, then workflow. What to hand over, what to gate, what to keep in your own hands.
Each person names the risk they had admitted and not changed, the correction they made this week, and what it cost. Then names the next one. This is the part a recording cannot replicate.
AI for Leaders & Business is a small, growing room of business owners and leaders who are putting AI to work without handing over the judgment that made them worth following. Your seat Wednesday is a membership, and the membership includes:
Full access from the moment you join, which means Wednesday night is inside your trial. Cancel anytime, no penalty. No upsells inside.
William is an ICF Professional Certified Coach and a retired U.S. Coast Guard officer with 29 years of active duty, who has coached executives at NASA, the CIA, the White House Leadership Development Program, the U.S. Army, and BAE Systems.
He is a Certified AI Consultant working at the intersection of AI literacy, executive identity, and business systems. His work teaches leaders how to multiply themselves before they try to multiply their output.
The premise behind everything he teaches: you cannot perform any greater than you think and feel at any given moment in time. AI does not fix you. It exposes you.
September 30 · 7 to 9 PM Eastern · Inside AI for Leaders & Business
Take the Seat →14-day trial. Join today and you are in the room Wednesday.