Why AGI is pure fantasy isn’t just a bold claim: the video walks through current AI capabilities and nails down five major hurdles—and realistic target dates—for reaching AGI. From mastering unified world models today, to seamless continuous learning by 2030, injecting true System 2 reasoning by 2035, crafting self-defined goals by 2040, and finally sorting out robust governance by around 2050, the talk shows we’re playing a long game. It even drops handy references like the LBM paper, JEPA, and Kahneman’s Thinking Fast and Slow.
If you’re losing sleep over an imminent superintelligence, this is your wake-up call—and your checklist. The hosts urge AI podcast guests and interviewers to sharpen their questions and avoid time-wasting tangents. Feeling inspired? There’s also a link to an AI Engineering course if you want to dive deeper.
Watch on YouTube
Top comments (0)