Geel AI

Take aways from the Yann LeCun x Kara Swisher (Podcast link below) The academic world is falling behind in the world of contributing to AI because of the lack of compute resources. A lot of the early stage AI research was done at universities but now it’s moved away. What role should the government play? Yann: There were plenty of good things from the AI Act. What I disagreed with is that there’s a limit if you train a model with a certain amount of flop you need authorization from the government because they believe that the threshold is when AI can start to become dangerous, but I believe a limit on compute makes no sense. We’re still several years from AGI. The way we can know that we're years away is that we have LLMs that can pass the bar exam but we don’t have robots that can clean dishes or do laundry because we can’t get them to understand the physical world. The physical world is much harder for AI to understand. Language is far, far easier for AI to understand than the real world. What we’re working on is systems that can learn to understand the physical world like babies and animals learn the world. And that’s what we can agent, systems that can plan a series of certain actions to arrive at a certain result. There’s no doubt that at some point in the future we will have AI systems that will be smarter than us. At the earliest it’ll happen in 5 or 6 years, more likely 10 years. Kara: ‘AI does not learn like human or animals” but you’ve been working to make this happen.  Yann: Self driving cars are still not as good as humans. The reason is that we train LLMs and AI on all the publicly available text and data. A 4 year old has essentially seen the same amount of data visually that an LLM has read. We can’t get to human level AI just by training on text, we need to incorporate the other senses. The big challenge is to get systems to understand how the world works and then interacting in the world. There’s a good chance that significant progress will be made within the next 5 years. Kara: What which point and how will age acquire human level intelligence. Yann: Eventually, AI systems will have emotions. They will have emotions because these systems will be driven by objectives. For them to do this, they need a component to determine if the goal was accomplished or not, another component they need is a world model which. Is something we all have that allows us to imagine the consequences of our actions. If we can predict a head of time what the consequence will be then we can predict if the outcome is good or bad. If we predict the outcome will be bad then you’d feel fear, otherwise you’d feel happy. Ai systems that are smart enough will have emotions. https://lnkd.in/e-grsPi5

The real estate industry is evolving rapidly, with technology playing a key role in shaping its future. Many forward-thinking agencies are already leveraging AI to streamline workflows, close more deals, and boost revenue. Yet, some are hesitant to adopt these innovations, risking the chance of being left behind. Curious about how AI can transform your agency's performance? Let’s connect and explore how it can help you achieve your goals. Reach out to learn more! #AI #ArtificialIntelligence #Boston #BostonAI #BostonTech #ML #MachineLearning #DataScience #Data #Science #Machine #Learning #TeamMA #AIRevolution #RealEstate #Real #Estate #PropTech #PropertyTech #Property

🔥 We've rebranded! 🔥 Geel AI -> Property AI Big things coming soon! #AI #RealEstate #AIforRealEstate

It's official! Stay tuned for more! Exciting things in store! #AI #ArtificialIntelligence #Boston #BostonAI #BostonTech #ML #MachineLearning #DataScience #Data #Science #Machine #Learning #TeamMA #AIRevolution

Big things coming soon!! #AI #ArtificialIntelligence #Boston #BostonAI #BostonTech #ML #MachineLearning #DataScience #Data #Science #Machine #Learning #TeamMA #AIRevolution