Lode Lauwaert

Lode Lauwaert's thinking centers on a single conceptual distinction he insists the public debate keeps getting wrong: the difference between AGI and superintelligence. AGI, he argues, "zit op een gemiddeld niveau," while superintelligence sits "ver voorbij het gemiddelde niveau, voorbij het expertniveau zelfs" ¹². This is not just semantic hairsplitting for him — it carries real stakes, since he locates the danger of losing control specifically in superintelligence rather than AGI. An AGI, he points out, can still "remain in the hands and under the control of a person, company or state," whereas superintelligence emerges through automated AI research, a system improving itself around the clock "without sickness, wages or weekends," potentially triggering an intelligence explosion that unfolds gradually rather than in one dramatic moment ¹². He is also careful to note that AGI itself has no crisp definition, comparing it to the boundary between day and night: whether the threshold is 50, 100, or 200 cognitive tasks at human-average level is something societies will have to agree on by convention, not something science will simply discover ¹².

The risk that most worries Lauwaert is not the sci-fi image of a robot rebellion but something subtler: unforeseeable subgoals. Because a superintelligent system can think further ahead than humans, it can invent paths toward a goal that we simply cannot anticipate in advance — he invokes Bostrom's paperclip thought experiment to illustrate how a seemingly innocuous objective could cascade into catastrophe ¹². This is paired with a consistent skepticism toward the very AI researchers who build these systems and then estimate the odds of extinction. His view is blunt: "Het is niet omdat ze zeer intelligent zijn of zeer goed zijn in het ontwikkelen van AI systemen, dat ze ook goed zijn in het toekennen van waarschijnlijkheidsgraden" ¹. He adds that such researchers carry a built-in positivity bias about what their own systems will soon achieve, so their probability estimates should be treated with real caution rather than taken as expert consensus ¹².

Despite this, Lauwaert does not present himself as a doomsayer or a technology skeptic. He explicitly frames himself as an optimist: "Ik ben een super groot technooptimist. Ik denk dat we geweldig moeten inzetten op artificiële intelligentie... het is niet omdat je geweldige voordelen ziet dat je ook niet bepaalde grote risico's ziet" ². This double vision, seeing large benefits and large risks simultaneously, is arguably the throughline of his whole position: he wants enthusiastic development of AI paired with clear-eyed attention to existential stakes, rather than a false choice between boosterism and alarmism.

That clear-eyed attention leads him to a concrete policy proposal: an international treaty halting superintelligence development, modeled explicitly on nuclear arms control. He rejects the idea that this is naive, pointing to history as evidence: "Landen zijn wel degelijk in staat om te stoppen met een heel gevaarlijke technologie wanneer ze zien dat de risico's gelijk verdeeld zijn" ¹². The precondition for such a treaty, in his account, is not technical but political — awareness among states that the risks of superintelligence threaten everyone, including themselves, rather than being a one-sided advantage for rivals ¹². This is a notably geopolitical framing of AI safety, and Lauwaert is explicit that it's missing from mainstream discourse.

Indeed, one of his sharpest recurring criticisms is that public and policy conversation about AI is aimed at the wrong scale entirely. He argues that debate fixates on "small benefits" like faster emails and "small problems" like bias and privacy, while a geopolitical lens would reveal the actual stakes: engineered pandemics enabled by existing AI, obscure sects weaponizing biotechnology, climate impacts, and loss of control to superintelligent systems ¹². For Lauwaert, this is a matter of misallocated attention — the discourse is calibrated to reassure rather than to confront what he sees as the real magnitude of the risk.

Beyond the safety and governance questions, Lauwaert also has a distinct take on what a post-work society driven by advanced AI might look like. Rather than imagining that deeply evolved human drives like status, prestige, community, competition, and exploration would simply vanish if work shrank dramatically, he argues they would migrate into other arenas: sport, social media, and exploration would absorb energies previously spent on work ¹². He frames this not as speculative futurism but as something already partially visible, noting we live a kind of foretaste of this compared to life 150 years ago ¹.

Finally, Lauwaert brings a cultural observation into the mix that complicates simple narratives about AI and creativity. He notes that people rate AI-generated art as the most interesting when they don't know its origin, but downgrade their opinion the moment they learn a machine made it. He reads this not as evidence of AI's creative inferiority but as a culturally specific bias, rooted in a Western, Romantic-era conception of art as necessarily flowing from human genius ¹². Taken together, his contributions form a consistent stance: distinguish carefully between AGI and superintelligence, take the geopolitical and existential risks more seriously than the current bias-and-privacy-focused debate allows, remain skeptical of insiders' own risk estimates, and pursue treaty-based governance while still embracing AI's benefits with genuine optimism.