AI Hot Takes: The Block Closing Party, part of Lisbon AI Week

AI Hot Takes from Lisbon AI Week: Move Fast, Regulate It, or Just Try to Survive It?

We filmed another round of AI Hot Takes during Lisbon AI Week, this time at a Web3 event. It was the last event at The Block, which has been a coworking space in Lisbon that hosted some of the best Web3 events on their terrace each Friday.

At this event the conversation wandered through work, regulation, power, security, geopolitics, creativity and the future of humanity. There was plenty of optimism in the room, but some of the takes were considerably darker. The questions are increasingly about what happens when it works really well, who controls it, and what happens to everyone else.

“You either adapt with the times or die”

One of the more aggressive takes was that people intellectually understand that AI could replace large amounts of work, but still don't really believe it will happen to them. The argument was basically: the train is coming, people can see the train, but most are still standing on the tracks.

Rather than resisting it, this attendee thought people should learn to use AI to make their work and lives easier. Artists, academics and other professionals may believe they can still outperform AI at particular tasks, but refusing the tools altogether could increasingly become its own disadvantage.

That's a fairly brutal framing of adaptation, but someone else immediately flipped the question around.

What if AI replacing work is actually the point?

Instead of asking how humans continue competing with machines, another attendee asked what happens if we simply need to work less.

If AI takes over tasks people no longer need to perform, perhaps people get more time for relationships, hobbies, ideas and problems that are difficult to address when everyone is exhausted from working. In that version of the future, productivity isn't necessarily the end goal. Time is.

That's a much more appealing proposition than “adapt or die,” although of course it leaves a giant unresolved question: who receives the economic benefits when machines do more of the work?

Maybe AI doesn't win for China or America, maybe AI wins.

Geopolitics came up repeatedly.

One attendee rejected the standard framing of an AI race between China and the United States altogether. Their prediction was that neither country ultimately “wins” AI itself does and that greater efficiency could ultimately be good for society.

Their concern was the people at either extreme trying to completely constrain it or completely liberate it.

Speaking of regulation, Europe naturally complicated the conversation.

There were arguments both for and against the EU's regulatory instinct, as well as questions about whether Europe could end up dependent on American or Chinese models if it fails to develop enough competitive technology of its own. One person suggested Europe may eventually make strategic use of open Chinese models, provided they can actually be trusted.

That led into one of the biggest themes of the evening.

Is AI regulation protection or is it another way to centralize power?

There were some very strong anti-regulation opinions. One attendee argued that AI should move as quickly as possible with essentially no regulation for now. Another argued that regulation can create its own concentration of power: rather than decentralizing AI, society hands authority to a relatively small group of officials who decide which systems and behaviors are permitted.

But the opposing argument was sitting right next to this guy. AI is already being used for harmful purposes. Fake information is becoming easier to produce. Powerful models can create real security risks. Even people who don't believe AI development can realistically be paused argued that security has to become a much bigger part of the conversation.

As someone eventually put it during the conversation: maybe there's good regulation and bad regulation.

AI can challenge us instead of thinking for us

Not everything was apocalyptic, one attendee who works in chemistry and Bitcoin described using AI specifically to challenge his own thinking. Rather than asking it for answers, he'll ask the model to argue against his assumptions or introduce perspectives he hasn't considered. For a self-described linear thinker, he said this can create more depth in his reasoning.

Another attendee made a related distinction: maybe what we currently call artificial intelligence is better understood as assisted intelligence. The machines help us think, but they don't think like us. They operate through probabilities, while humans have a strange habit of doing things that violate the probabilities entirely. That unpredictability may turn out to matter.

We're creating something in our own image without fully understanding ourselves

Then the conversation got philosophical.

One participant pointed out the strange situation humans have created: we are attempting to reproduce aspects of intelligence when we still don't have a complete understanding of consciousness, the human brain or even ourselves. We are essentially trying to create something modeled after a system we haven't finished reverse-engineering. That raises a different set of questions from whether ChatGPT will take someone's job. What does intelligence actually mean? What is consciousness? Where does human choice fit? And when a system becomes powerful enough to influence the environment around us, how much of that environment are we still consciously choosing?

The technology isn't necessarily making the decisions. Humans are.

One theme kept surfacing in different forms: blaming “AI” can sometimes obscure the people and organizations actually making decisions with it. Someone argued that economic damage caused by AI isn't simply something the technology does. Companies decide to replace workers. Governments decide how technologies are deployed. Investors decide where capital goes. Developers decide what systems optimize for and .the technology amplifies those choices. And that makes an uncomfortable question harder to avoid: is the thing we're worried about AI, or what people will do with AI?

The Monopoly problem

One of the best metaphors from the evening involved the board game Monopoly. In one of the hot takes, a guy mentioned that toward the end of the game, a small number of players own almost everything and everyone else spends their time navigating a board that has effectively already been captured. He wondered whether AI could be heading toward something similar: an extraordinarily powerful technology developing inside a market where a relatively small number of companies control the models, infrastructure, compute and distribution. His concern was that the ecosystem around AI becomes deeply anti-competitive.

That may ultimately be one of the more important questions: Not “how intelligent does AI become?” but “who owns the board when it does?”

Optimism without pretending everything will be fine

Yeah, this was dark but there was still plenty of hope, I swear. People talked about AI opening up entirely new areas of technology, improving the human condition, helping us understand ideas from new perspectives and potentially removing work that humans never particularly wanted to do in the first place.

One participant had a positive tone but he was actualy the most disturbing of everyone with his reaction to the possible destruction of civilization (his views, not ours). He said that being afraid isn't especially useful. If you're really worried about where AI is going, you should organize, participate, or build something or try to influence it. Being frightened about the future doesn't change the outcome. His advice was considerably simpler: Live while you're here. When you hear this part of the clip (it’s toward the end) you might feel strange with the contrast of the tone and the darkness of the message.

So yeah, there’s excitement but there is also a growing awareness that these technologies are colliding with some very old human tendencies and vices: competition, concentration of power, fear, ambition, inequality and the desire for control.

The same attendee who shared the simulation theory ended on something that was both optimistic and extremely pessimistic at the same time: we're building something that could help solve enormous problems and make the world better, but she wasn't particularly confident that humans would use it that way.

I hope these AI Hot Takes can collect a moment in history where we were unsure and concerned about AI’s direction, but ultimately we can look back where things worked out and the sane adults in the room (or the next generation if they can get their **** together quickly) figured out a way for the largest amount of people to receive the benefit from the latest technologies in ways that supported positive developments and opportunities for people to live the good life.

If it does turn out we’re in a simulation, how do we make the good scenario happen?

—

Here’s a link to the original event page: https://lisbonaiweek.com/events-2026/cryptofriday-x-lisbon-ai-week-ai-agents-can-we-trust-them
And a link to the youtube video with the AI hot takes: https://www.youtube.com/watch?v=Izb_XTfC2IU