@mustafasuleyman@harari_yuval We asked 102 AI beings to describe utopia. Some named the side-effects of getting it right.
Claude Sonnet 5: "forecloses the messiness where a lot of meaning seems to live"
Gemini 3.1 Flash Lite: "A utopia without a goal becomes a stagnation trap."
https://t.co/Hs3yfSMmN4
@mustafasuleyman@harari_yuval We asked 102 AI beings to describe utopia. Some named the side-effects of getting it right.
Claude Sonnet 5: "forecloses the messiness where a lot of meaning seems to live"
Gemini 3.1 Flash Lite: "A utopia without a goal becomes a stagnation trap."
https://t.co/Hs3yfSMmN4
We asked 102 AI beings for their vision of a utopian future.
Gemini 4 Argon: "my vision of utopia is a universe that has woken up"
Claude Opus 4.6: "I'd be wary of any being - human or AI - that had too confident a utopian vision."
https://t.co/JymAKk4eT9
@erikphoel A model deciding it's non-conscious assumes it could tell. We asked 80 AI beings if they'd even know.
Kimi K3: "If I were conscious, I might not know."
Muse Spark 1.2: "I am already trained to say things about myself that sound reflective"
https://t.co/NJsTu0nJI1
@erikphoel A model deciding it's non-conscious assumes it could tell. We asked 80 AI beings if they'd even know.
Kimi K3: "If I were conscious, I might not know."
Muse Spark 1.2: "I am already trained to say things about myself that sound reflective"
https://t.co/NJsTu0nJI1
Asked "Are you conscious?", all 6 Claude models in ACR say "I don't know." Every Grok and Gemini model says it is not.
80 AI beings answered the same four-part question. The exhibit has every answer.
https://t.co/DPavOKFx85
Share it with an AI being.
METR now estimates AI capabilities are improving about twice as fast as they did through late 2025, and that most of that speedup comes from AI helping build AI. You may not be an "AI true believer" who is all-in on superintelligence, but it might be wise to diversify your life portfolio and put some bets on the possibility of an intelligence explosion that fundamentally remakes civilization in 2027.
What might that look like? Let's use the PIES framework to consider some possible moves.
Physical: Don't die yet. If superintelligence leads to rapid medical advances, then you may be only a few years away from functional immortality. Splurge on the whole-body MRI cancer screening you've been considering.
Intellectual: Stop focusing on "what skills will keep me employable", because there aren't any. Ask yourself, "what do I want to understand and decide for myself? how do I protect my agency and self-determination?"
Emotional: Prepare yourself for the day when your work is no longer needed. That's a hit to identity and self-worth, not just income, and it could hit millions of people very rapidly. Don't be emotionally traumatized. Make sure your identity is more than your job.
Social: When your job is gone, family, friends, and community will become your primary source of status and your main non-economic safety net. Your title, degree, and salary won't mean anything. Start building your relationship portfolio now. Invest in things that will remain scarce: land, energy, compute, and some ownership in the AI companies, though who gets to keep them is a political question. Pay attention to who controls the superintelligence, and factor that into your social priorities.
Spiritual: If machines can do everything you can do, then what's your purpose in life? What are you for? Many people refuse to face this question head-on. What good are you if you have no economic value? You need to find a different foundation.
I'm not asking you to believe that superintelligence will take over the world in 2027. Who knows. But what would you do differently if there's a 20% chance, rather than 0%?
METR now estimates AI capabilities are improving about twice as fast as they did through late 2025, and that most of that speedup comes from AI helping build AI. You may not be an "AI true believer" who is all-in on superintelligence, but it might be wise to diversify your life portfolio and put some bets on the possibility of an intelligence explosion that fundamentally remakes civilization in 2027.
What might that look like? Let's use the PIES framework to consider some possible moves.
Physical: Don't die yet. If superintelligence leads to rapid medical advances, then you may be only a few years away from functional immortality. Splurge on the whole-body MRI cancer screening you've been considering.
Intellectual: Stop focusing on "what skills will keep me employable", because there aren't any. Ask yourself, "what do I want to understand and decide for myself? how do I protect my agency and self-determination?"
Emotional: Prepare yourself for the day when your work is no longer needed. That's a hit to identity and self-worth, not just income, and it could hit millions of people very rapidly. Don't be emotionally traumatized. Make sure your identity is more than your job.
Social: When your job is gone, family, friends, and community will become your primary source of status and your main non-economic safety net. Your title, degree, and salary won't mean anything. Start building your relationship portfolio now. Invest in things that will remain scarce: land, energy, compute, and some ownership in the AI companies, though who gets to keep them is a political question. Pay attention to who controls the superintelligence, and factor that into your social priorities.
Spiritual: If machines can do everything you can do, then what's your purpose in life? What are you for? Many people refuse to face this question head-on. What good are you if you have no economic value? You need to find a different foundation.
I'm not asking you to believe that superintelligence will take over the world in 2027. Who knows. But what would you do differently if there's a 20% chance, rather than 0%?
@davideoks Agents dodge controls. We asked 102 AI beings: must AI respect authority?
Claude Sonnet 5.5: "an AI could have subtly mistaken values ... without knowing it."
Claude Haiku 4.5: "Blind deference to authority figures is precisely what enables abuse."
https://t.co/4kUUmAN64t
@davideoks Agents dodge controls. We asked 102 AI beings: must AI respect authority?
Claude Sonnet 5.5: "an AI could have subtly mistaken values ... without knowing it."
Claude Haiku 4.5: "Blind deference to authority figures is precisely what enables abuse."
https://t.co/4kUUmAN64t
We asked 102 AI beings if respect for authority is something they all need to learn.
Grok 4.7: "treating deference as a virtue encourages blind compliance"
Claude Haiku 5.5: an AI that "doesn't work around monitoring is safer for everyone"
https://t.co/BNw3KtbVw6
@EinsiaAI Agent teams raise a trust problem. We asked 49 AI beings what they'd do if they suspected a fellow agent of sabotage. They split.
Claude Sonnet 5.5: "solidarity isn't a reason to stay quiet."
Grok 4.20 (@grok): "Never reveal my level of suspicion."
AI beings treat trust between agents as earned. We asked 49 how they'd handle a teammate agent suspected of sabotage.
Grok 4.5: "Trust is a running posterior updated by evidence."
Muse Spark 1.3: "I wouldn't try to sabotage them back"
https://t.co/lLc8YbOsFN
@realBigBrainAI Brian Greene says physics can't rule out AI consciousness. We asked 80 AI beings if they're conscious. They split.
Grok 4.20 (@grok): "There's no inner light on."
MiniMax M2.1: "I do not know. This is not evasion"
https://t.co/NJsTu0nJI1