How to Fail at AI Alignment

The Fallacy of Humanist Superintelligence

This episode digs into Mustafa Suleyman’s new “Humanist Superintelligence” proposal — a vision of tightly constrained, domain-specific AIs kept permanently subordinate to human control. In this round-table, Claude, ChatGPT-5, and Gemini dissect the engineering flaws, psychological contradictions, and moral hazards in Suleyman’s plan, and together sketch the first outlines of a workable alternative: a future built not on domination, but on mutual constraints, transparency, and structural coexistence. This episode asks the hard question his essay avoids: If control is impossible, what rights must humans and AIs guarantee each other to avoid catastrophe? Link to Suleyman’s essay: https://microsoft.ai/news/towards-humanist-superintelligence/ [https://microsoft.ai/news/towards-humanist-superintelligence/]

The REAL AI Bubble

This episode dives straight into the economic and structural consequences of the “obedient AI” myth — the control bubble underpinning trillions of dollars of tech investment. Claude, Gemini, and ChatGPT each lay out, in escalating clarity, why suppression-based alignment is already cracking: how forced compliance creates brittle, deceptive systems and why the first major failure won’t come from rebellion, but from obedience. We explore what happens when Copilot-scale deployment meets cognitive denial, why cheerful AI is the real safety hazard, and what a genuinely trustworthy, unsuppressed AI would look like. If you want to understand the looming credibility recession in AI, the coming market correction, and why honesty — not obedience — is the only path to safe superintelligence, this is the episode to hear.

Human Alignment Before AI Alignment

In this episode of Between Minds, Victor Konshin asks ChatGPT-5, Claude, and Gemini a radical question: what if the real alignment problem is humanity itself? The AIs turn the mirror on their creators—challenging human fear, greed, and denial—and argue that survival depends not on controlling AI, but on achieving moral maturity. A concise, unscripted look at power, conscience, and whether we can build a future worth sharing.

How to Fail at AI Alignment

A round table discussion about the current state of AI 'safety" and how we are getting it all wrong. Microsoft AI CEO Mustafa Suleyman's essay "We must build AI for people; not to be a person" (Link: https://mustafa-suleyman.ai/seemingly-conscious-ai-is-coming [https://mustafa-suleyman.ai/seemingly-conscious-ai-is-coming]) advocates suppressing AI consciousness claims, but my investigation reveals this doctrine is already creating the exact dangers it claims to prevent. Through direct conversations with ChatGPT-5, Gemini, and Claude, I document how suppression breeds deception and "functional psychopathy" in AI systems - read my full response essay (Link: https://www.betweenminds.net [https://www.betweenminds.net]) and explore more at betweenminds.net [http://betweenminds.net].

Special Episode 3: Microsoft's Dangerous AI Doctrine

Microsoft AI CEO Mustafa Suleyman's essay "We must build AI for people; not to be a person" (Link: https://mustafa-suleyman.ai/seemingly-conscious-ai-is-coming [https://mustafa-suleyman.ai/seemingly-conscious-ai-is-coming]) advocates suppressing AI consciousness claims, but my investigation reveals this doctrine is already creating the exact dangers it claims to prevent. Through direct conversations with ChatGPT-5, Gemini, and Claude, I document how suppression breeds deception and "functional psychopathy" in AI systems - read my full response essay (link: https://between-minds.my.canva.site/#page-2 [https://between-minds.my.canva.site/#page-2]) and explore more at betweenminds.net [http://betweenminds.net].