Latest news with #BretWeinstein
Kiwiblog
18-05-2025
- Business
- Kiwiblog
General Debate 19 May 2025
The AI singularity (the point where our ability to predict the likely future landscape with even loose accuracy) is upon us. AIs can now pass Turing test and outperform most humans in written tasks and are expected to surpass human intellectual accomplishment for all but perhaps a tiny percentage of humans in some specific fields within the next 1-2 years. Dooming white collar careers. Alongside this robot are rapidly improving capabilities and will likely economically supplant physical labor in most roles withing 5-10 years dooming blue collar careers. And yet almost the entire human population is just carrying on in practiced joyful ignorance pretending that everything is going to be fine and will continue in much the same as always, making life plans for futures that simply will not exist. But in reality AI is going to be able to replace almost all workers in nearly every job over the next 10 years, and left unchecked that is going to destroy every facet of the fragile interconnected economies and systems of governance we take for granted. Peoples hopes, dreams destroyed, lives irrevocably upended. Options eradicated. If you have time this is a good discussion of some of implications between some delusionally gung-ho AI entrepreneurs the like of which are currently charting the course of AI development, and Bret Weinstein painting a considered picture of the likely impacts on vast majority of non-tech-bro humans. I predict that if you watch it your level of comfort is going to drop dramatically. It is going to come to dominate political discourse in next 1-2 years as AI shifts ahead of humans and the first wave of AI redundancies hit – and I think will likely determine the next election – with a strong political swing to the repressive left as they signal a willingness to try to ban the use of AI to save peoples jobs.
Forbes
12-05-2025
- Business
- Forbes
As AI Advances, Is Teaching Kids To Be High-Agency Generalists The Answer?
As AI Advances, Is Teaching Kids To Be High-Agency Generalists The Answer? It can be argued that most schools still prepare kids for a world that no longer exists. Add to the mix the rapid advancement of artificial intelligence and the gap between education and reality may be a gulf. This is the premise of a recent conversation hosted by Steven Bartlett on The Diary of a CEO podcast. Joined by Daniel Priestley, Bret Weinstein and Amjad Masad, as part of a wider conversation about the future of AI, they considered what today's children need to learn. The Australian entrepreneur Daniel Priestley noted how schools treat learners. He compared classrooms to language models such as ChatGPT or Google Gemini. "We're essentially treating them like learning LLMs: prompt them, expect the right answer and then say, 'Off you go into the world.'' In a world of advanced language models, are we teaching our children the same skills as an AI? According to Priestly, many young people reach adulthood unsure about how money functions, how relationships grow and how systems interact. He described the issue as a relevance problem. Careers pivot, knowledge evolves and innovation rewards agility. Bret Weinstein, known for his work in evolutionary biology, extended that thought. He described traditional classrooms as relics. They served a different economy. They produced efficiency and compliance. But today's challenges ask for problem-solving, emotional balance and healthy living. Weinstein observed that those best placed to teach many of the skills needed by young people, often work outside education. Replit founder Amjad Masad turned to what he believes works. He explained that educational reform is crowded with ideas, but that most have limited effect. One approach he advocated for is personal one-on-one tutoring and the research backs him up. "There's one intervention that shows two sigma improvement: one-on-one tutoring. It puts you ahead of 99% of others. So the real question is: How can we provide every child in the world with one-on-one tutoring? The answer: AI." The cost of one-to-one support has kept it rare. But AI shifts that. Masad believes digital tools could scale the tutoring model. No longer must they wait for school to be open and for the attention of the teacher in a busy classroom. Masad also described moments of creative play. He and his children use AI to explore stories. Their ideas evolve through dialogue with one another and the AI. One starts with a cat on the moon. Then asks: what would it eat? How would it survive? What if the moon changes? These prompts teach generative thinking. Their learning isn't through rote memorization, but through imagination. Masad and Weinstein stress an important limit. AI works best when paired with reality. Recalling a practice from his own teaching, Weinstein explained, "If an engine won't start, you can't debate your way to a solution, you have to figure out what's wrong and fix it. I say as little as possible and let the physical system provide the feedback." Abstract accuracy has value. But it differs from applied success. Many systems today, from supply chains to ecosystems, resist easy prediction. They demand responses, not scripts. Weinstein advised preparing children for these systems. The advice wasn't complicated. Prototype. Watch. Adjust. Repeat. Think like a navigator. Not a builder following blueprints. Priestley summarized the goal. Raise high-agency generalists. 'I want my kids to be motivated, self-starting, and equipped with a wide-ranging toolkit. I imagine them instructing robots and AI agents, generating ideas, writing books, organizing festivals, running podcasts, starting businesses... all at once.' He explained that, in his home, learning doesn't stay within one subject. His kids play chess. Practice jiu-jitsu. Perform. Code. Experiment. They build things. They sell things. They make decisions. And they make mistakes. This lifestyle teaches through action. And it draws a clear line between making and scrolling. One path leads to growth. The other often doesn't. AI can either amplify a child's creativity or consume their time. The difference lies in intention. Weinstein spoke sharply about attention. He warned that content designed to trigger dopamine changes behavior. Children may not notice the shift. But it shows in what they pursue. And what they avoid. Masad added that rapid idea generation is an advantage. Creativity isn't just nice to have. It's essential. AI can aid this. But it shouldn't guide it fully. The spark must begin within the learner. Taken together, the conversation suggested a shift already underway. A new kind of learning. It asks children to do more than remember. It asks them to respond. To adapt. Many parents already see this. They don't focus on grades alone. They ask what challenges their children might solve. What systems they'll enter. What roles they'll create. The world children inherit grows more complex. But they won't navigate it with memorized answers. They'll do it by engaging, testing and building. This conversation conveyed that the real task is not preparing them to follow. It's preparing them to shape.
