Jul 15, 2026 · 1h 0m · big-technology
AI Pioneer Jürgen Schmidhuber: AI Already Feels Pain, Loves, and Is Self-Aware
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the Big Technology Podcast, AI pioneer Jürgen Schmidhuber explores the civilizational roots, computational mechanics, and economic realities of artificial intelligence, arguing that true AGI requires embodied world models, machine consciousness, and decentralized access rather than language-only paradigms.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
Alex as informed peer 5.5 Guest teaching 6.4 Guest disagreement 4.5 Alex pushing back 3.4
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 21:00 Schmidhuber attacks AI economic sustainability
Schmidhuber aggressively rejects the conventional belief that Big Tech is winning AI, forecasting 900 billion dollar losses and calling out unsustainable zombie unicorns and lack of business models.
Hardest push from Alex ▶ 34:11 Kantrowicz challenges AI emotions with biological neurochemistryKantrowicz firmly refuses Schmidhuber's functionalist definition of pain and love, pointing out that robots lack physiological systems and neurochemicals like oxytocin.
Biggest teaching moment ▶ 8:40 Schmidhuber contrasts LLM scraping with baby-like active learningSchmidhuber provides a detailed technical explanation contrasting passive LLM token prediction with active world-model experimentation performed by infants and physicists.
Alex holds their own ▶ 7:34 Kantrowicz cites Greg Brockman inside OpenAI on model strategyKantrowicz demonstrates industry access and analytical depth by challenging Schmidhuber with Greg Brockman's insider rationale for pausing Sora to double down on GPT architectures.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| Deconstructing the 'Father of AI' and Civilizational AI Development | 4 | 6 | 4 | 2 | Kantrowicz introduces Schmidhuber by citing media titles like 'father of AI', but Schmidhuber immediately rejects the individualistic framing, arguing that creating AI requires an entire civilization. Schmidhuber then articulates why Turing tests are a flawed measure of true intelligence compared to physical mastery like plumbing. | |
| World Models vs. LLMs: Why Language Models Cannot Achieve AGI | 7 | 7 | 6 | 5 | Kantrowicz leverages his firsthand conversation with OpenAI President Greg Brockman regarding the pause on Sora to question whether LLMs alone can reach AGI. Schmidhuber firmly rejects the premise that LLMs suffice, explaining that true AGI requires active world models conducting experimental loops rather than passive web downloads. | |
| The Hardware Bottleneck in Physical Robotics and Humanoid Dexterity | 5 | 6 | 3 | 2 | Kantrowicz quotes Schmidhuber's 2017 prediction about dexterous robotics replacing human labor. Schmidhuber explains why robotic physical hardware has lagged drastically behind exponential compute gains since 1941, highlighting the complexity and self-healing nature of the human hand. | |
| The AI Economic Bubble and Value Capture for the 'Little Man' | 6 | 6 | 6 | 4 | Kantrowicz challenges Schmidhuber's optimistic claim that the 'little man' captures the most value from AI, noting current concentration in trillion-dollar tech giants. Schmidhuber forcefully counters that big tech CapEx spending creates massive cash flow deficits and unviable utility-like debt traps destined for a bubble burst. | |
| Sponsorship Break: Gravity AI Agent Security Documentary | 6 | 7 | 5 | 6 | Kantrowicz pushes back on the claim that AI experiences pain and love by citing biological neurochemistry like oxytocin and genuine conscious perception. Schmidhuber dismisses the distinction as superficial, arguing that biological reward-minimizing circuits and artificial objective functions are computationally equivalent. | |
| Machine Consciousness, Self-Agency, and Dual-Network Architectures | 6 | 7 | 5 | 3 | Kantrowicz brings up Geoffrey Hinton and the 'stochastic parrot' critique of LLM consciousness. Schmidhuber agrees LLMs merely mimic text, but details his historical 1991 dual-network architectures (conscious chunker and subconscious automatizer) to argue self-awareness has long existed in predictive control systems. | |
| Mind Uploading, Cosmic AI, and the Fate of Biological Humanity | 5 | 5 | 3 | 2 | Kantrowicz references Schmidhuber's past quotes on living in a simulation and asks if humanity must merge with AI to remain relevant. Schmidhuber elaborates on mind uploading, positing that biological humans will either expand cognition massively or remain irrelevant nostalgic artifacts in a native AI cosmos. | |
| Algorithmic Physics, Computable Universes, and Determinism | 5 | 7 | 4 | 3 | Kantrowicz asks whether deterministic AI models eliminate human free will. Schmidhuber details his algorithmic universe theory, arguing that physical reality is completely computable and that human concepts of free will are illusory. |
Statements from this episode (14)
Schmidhuber says AI development requires an entire civilization's contributions
“No single person can create an AI by himself or herself. You need an entire civilization to build an AI.”
Schmidhuber: Turing test is a bad metric for measuring true intelligence
“However, it just means that the Turing test is a bad way of measuring intelligence, because there is no AI in the physical world, outside of the screen, that can do all the things that a little boy can do, that can do the things that a plumber can do.”
Schmidhuber: Robots will soon learn to operate all existing physical machines
“And I think the culmination point of that, which I've been talking about again for decades, is at some point in the not so distant future, we will have a robot that doesn't have to be super smart, but just smart enough to learn to operate all the already exist…”
Schmidhuber: Large language models alone cannot lead to AGI
“Yeah, so when the phrase, when the question is phrased like this, does an LLM, a large language model by itself, lead to AGI, the answer is a clear no.”
Schmidhuber: AGI requires a separate controller to plan with world models
“The model alone, the foundation model alone, is, is not an AGI. No, you need the other guy, which uses all kinds of tricks to exploit The algorithmic information in the world model to come up with better plans.”
Schmidhuber: Physical robotics hardware evolves much slower than compute per dollar
“The artificial hands, the artificial bodies that we are trying to build for humanoid robots and so on, they are evolving much less rapidly than the compute per dollar. And I think that's the main problem.”
Schmidhuber: Trillion-Dollar AI GPU CapEx Will Lose $900B Within Five Years
“Now if you invest one thousand billion dollars today into GPUs for data centers, this means that within five years you are going to lose nine hundred billion dollars, you know. Somebody is going to lose nine hundred billion dollars in the near future because t…”
Schmidhuber: Proprietary AI Labs Have No Moat Against Open Source
“None of the companies have has a mode, you know, because whenever there's a new benchmark breaking record or something, benchmark, record breaking language model that does this or this or whatever. A few months later, there's the same thing in open source, you…”
Schmidhuber: AI Will Soon Run Locally Without Internet or Cloud
“Soon AI will not be in the cloud, you know. No, it will be local, on your local computer, a small little computer, it will be as powerful as what's now in the cloud, but you won't have to connect it to the internet”
Schmidhuber: AIs Have Felt Pain Since at Least 1990
“I always think it's funny that many people claim, even computer scientists claim, that AIs cannot feel pain because our AIs have felt pain At least since 1990, you know.”
Schmidhuber: No Objective Distinction Exists Between Biological and Artificial Emotions
“There's no way of objectively seeing the difference between the emotions of a Of reward maximizing biological brain and the emotions of a reward maximizing artificial brain.”
Schmidhuber: AI Achieves Self-Awareness When World Models Simulate Self-Representation
“Whenever the agent is, is making plans like that, it's thinking about itself, waking up these internal representations of itself, and then it's self-aware.”
Schmidhuber: No physical laws prevent uploading human minds into computers
“There is no physical reason to reject the notion that this might be possible.”
Schmidhuber: Future universe-shaping decisions will be made by AIs, not humans
“So either you are nostalgic and remain irrelevant, or you become part of this growing sociology, if you will, of AIs, which are mostly AIs, and some of them may have human roots and whatever, but almost all the decision-making process and the universe-shaping …”