A Robotics Pioneer's Surprising Predictions About AI and Humanoids

Rodney Brooks, a renowned MIT roboticist and pioneer, believes that people are vastly overestimating the capabilities of generative AI. Here's why:

1. Generative AI is not human-like: Brooks argues that generative AI systems like ChatGPT are not human and not even humanlike. It's flawed to try and assign human capabilities to them. People tend to overestimate the competence of these systems based on their performance on specific tasks.

2. Generative AI has limitations: While generative AI is capable of performing certain tasks, it cannot do everything a human can. Brooks says that people often generalize the capabilities of these systems beyond their actual competence.

3. Practical applications may not make sense: Brooks offers the example of using a large language model to control warehouse robots. In his view, this would be an inefficient and impractical use of generative AI, as it would slow down the system. Instead, it's much simpler to connect the robots directly to the warehouse management software.

4. Exponential growth is not guaranteed: Brooks challenges the belief that technology will always grow exponentially, as suggested by Moore's Law. He uses the example of the iPod, where storage capacity did not continue to double indefinitely. Similarly, he believes that the capabilities of language models may not grow as exponentially as some predict.

5. Humanoid robots face significant challenges: Brooks has extensive experience in building humanoid robots and believes that the common belief in their near-term potential is misguided. He predicts that it will take at least another 25 years before humanoid robots play a significant role, contrary to the claims of some entrepreneurs.

In summary, Rodney Brooks provides a more cautious and grounded perspective on the current state and future potential of generative AI and robotics. He cautions against the hype and overconfidence surrounding these technologies, emphasizing the need for a more realistic assessment of their capabilities and limitations.

Rodney Brooks, a renowned MIT roboticist and pioneer, believes that people are vastly overestimating the capabilities of generative AI. While he acknowledges the impressive nature of these technologies, he cautions against overestimating their abilities.

Brooks explains that the trouble with generative AI is that it can perform a certain set of tasks well, but it cannot do everything a human can. Humans tend to generalize the capabilities of AI systems based on their performance on specific tasks, often being overly optimistic about their overall competence.

He emphasizes that generative AI is not human and not even humanlike, and it is flawed to try to assign human capabilities to it. People see it as so able that they want to use it for applications that don't make sense, such as using large language models to control warehouse robots, which would actually slow things down.

Brooks also challenges the belief that technology will always grow exponentially, as seen with Moore's law. He uses the example of the iPod, where storage capacity did not continue to double indefinitely, as many had expected. Similarly, he believes that the exponential growth in the capabilities of language models may not continue indefinitely.

While Brooks acknowledges that large language models could potentially help with specific tasks, such as assisting with domestic robots for an aging population, he cautions that even this comes with its own unique set of challenges. The problem, he says, is not about being able to do the tasks, but rather about the underlying control theory and optimization required.

In summary, Rodney Brooks, a respected voice in the field of robotics and AI, urges caution in the face of the hype surrounding generative AI. He believes that these systems have limitations and that it is important to maintain a realistic perspective on their capabilities and potential applications.

Rodney Brooks, a renowned MIT roboticist and pioneer, believes that the hype around humanoid robots is vastly overestimated. He has learned from his decades of experience in the field that the form factor of humanoid robots is not the most effective solution for practical applications.

Brooks explains that the key is to make the technology accessible and purpose-built, rather than focusing on creating human-like robots. He uses the example of his current company, Robust.AI, where the robots resemble shopping carts with handlebars. This design allows for easy intervention by humans if there are any issues with the robot.

According to Brooks, the trouble with humanoid robots is that they are not human and not even human-like. Trying to assign human capabilities to them is a flawed approach. He has found that it's much more effective to design robots that can work alongside humans, rather than attempting to replicate human form and function.

Brooks also cautions against the misconception that technology will always grow exponentially, as suggested by Moore's Law. He uses the example of the iPod to illustrate that technology doesn't always follow a linear trajectory. The same principle applies to the development of large language models (LLMs) and AI systems.

While Brooks acknowledges that LLMs could potentially assist with specific tasks, such as in the context of an aging population, he emphasizes that the real challenges lie in control theory and other fundamental mathematical optimizations, rather than simply replicating human capabilities.

In summary, Rodney Brooks' perspective challenges the prevailing hype around humanoid robots and suggests a more pragmatic approach focused on designing purpose-built, accessible robotic solutions that can effectively collaborate with humans.

Brooks acknowledges that there is a mistaken belief, mostly thanks to Moore's law, that there will always be exponential growth when it comes to technology. The idea that if ChatGPT 4 is this good, imagine what ChatGPT 5, 6, and 7 will be like.

He says this is flawed logic and that technology doesn't always grow exponentially, despite Moore's law. He uses the example of the iPod - for the first few iterations, it did in fact double in storage size from 10 to 160GB. If it had continued on that trajectory, he figured we would have an iPod with 160TB by 2017. But of course, that didn't happen. The models being sold in 2017 actually came with 256GB or 160GB because, as he pointed out, nobody actually needed more than that.

While this is a valid point that Moore's law doesn't always apply to every analysis of technology, Brooks' iPod example may not be the most applicable comparison. A more relevant comparison would be to look at processor speed, as that is more akin to the computing power and capabilities of language models like ChatGPT.

Nevertheless, Brooks' main argument is that we should be cautious about extrapolating exponential growth indefinitely when it comes to technology. Just because a system demonstrates impressive capabilities today, it doesn't necessarily mean those capabilities will continue to grow at the same rate in the future. Real-world constraints and practical needs often temper the pace of technological advancement.

Rodney Brooks acknowledges that large language models (LLMs) could potentially help with domestic robots, especially in assisting an aging population where there are not enough people to provide care. However, he cautions that this could come with its own unique set of challenges.

Brooks explains that the problem is not about being able to do the tasks, but rather about the control theory and other hardcore math optimization required. He states that "people say oh the large language models are going to make robots do things they couldn't do but that's not the where the problem is."

While LLMs may be able to help with specific tasks for domestic robots, Brooks believes there are still significant technical hurdles to overcome. He emphasizes that the fundamental research breakthroughs required for practical domestic robots are not trivial, despite the optimism of some naive entrepreneurs.

Overall, Brooks takes a measured and cautious approach to the potential applications of generative AI in domestic robots. He acknowledges the possible benefits, but also highlights the substantial challenges that must be addressed before such systems can be widely deployed and effective.

Rodney Brooks, a renowned MIT roboticist and pioneer, offers a cautionary tale about the current state of AI and the startup ecosystem. He emphasizes the need for rational thinking amidst the hype and overestimation surrounding generative AI.

Brooks acknowledges the impressive capabilities of large language models (LLMs) like ChatGPT, but cautions against overestimating their abilities. He explains that while these systems can perform certain tasks well, they are not human and lack the full range of human capabilities. This tendency to generalize the performance of an AI system on a specific task to broader competence is a common pitfall.

Regarding humanoid robots, Brooks has a unique perspective. Despite having built and delivered more humanoid robots than anyone else, he believes the current approach is misguided. He advocates for practical, purpose-built robots that prioritize accessibility and ease of use, rather than humanoid forms.

Brooks also addresses the concerning trend of startup fraud and overinflated claims in the tech industry. He cites examples of CEOs in Silicon Valley who have faced legal consequences for misleading investors and the public. This pattern, he argues, is a result of the "fake it till you make it" mentality that has become prevalent in the startup culture.

The author's blog, where he maintains a scorecard of his predictions, provides valuable insights into his rational approach to technological advancements. His predictions, which have been largely accurate, suggest a measured and evidence-based perspective on the future of AI and robotics.

In conclusion, Rodney Brooks' insights serve as a reminder to approach the hype surrounding AI and startups with a critical and rational mindset. As the technology landscape continues to evolve, it is essential to separate fact from fiction and maintain a balanced understanding of the capabilities and limitations of these emerging technologies.

Rodney Brooks, a renowned MIT roboticist and pioneer, believes that people are vastly overestimating the capabilities of generative AI. While he acknowledges the impressive nature of large language models (LLMs) like ChatGPT, he cautions against overconfidence in their abilities.

Brooks explains that the trouble with generative AI is that while it can perform certain tasks well, it cannot do everything a human can. Humans tend to overestimate the competence of AI systems, generalizing their performance on specific tasks to a broader range of capabilities. However, Brooks emphasizes that generative AI is not human and not even humanlike, and it is flawed to try to assign human-like abilities to it.

He offers the example of his own company, Robust.ai, where someone suggested using an LLM to control the warehouse robots. Brooks believes this would not be a reasonable use case for generative AI and would actually slow things down. Instead, he prefers to connect the robots directly to the warehouse management software, which is a simpler and more effective solution.

Brooks also challenges the common belief that technology will always grow exponentially, as suggested by Moore's law. He uses the example of the iPod, where storage capacity did not continue to double indefinitely, as many had expected. Similarly, he believes that the exponential growth in the capabilities of LLMs may not be a reliable predictor of future progress.

While Brooks acknowledges that LLMs could potentially help with domestic robots, especially in caring for an aging population, he cautions that this too could come with its own unique challenges. He emphasizes that the problem is not just about being able to do tasks, but also about control theory and other complex mathematical optimizations.

Overall, Brooks' perspective provides a more nuanced and cautious view on the current state and future potential of generative AI. His decades of experience in the field of robotics and AI research give his insights significant weight, and his predictions about the emergence of the "next big thing" in AI, which he believes to be neuro-symbolic AI, are worth considering as the field continues to evolve.

Rodney Brooks, a renowned MIT roboticist and pioneer, has made several predictions about the future of robotics and AI. Here are some of his key insights:

Generative AI Hype

Brooks believes that people are vastly overestimating the capabilities of generative AI. While impressive, these systems have limitations and cannot do everything a human can. He cautions against assigning human-like capabilities to them.

Humanoid Robots

Brooks is skeptical about the near-term potential of humanoid robots. Based on his extensive experience, he predicts that it will take at least another 25 years before humanoid robots play a significant role. He prefers practical, purpose-built robots over humanoid designs.

Exponential Growth in Technology

Brooks challenges the notion of exponential growth in technology, using the example of the iPod to illustrate how growth can plateau. He argues that the same principle applies to the development of large language models and other AI systems.

Robotics and Aging Population

Brooks acknowledges that large language models could potentially assist with tasks for an aging population, but cautions that the real challenges lie in control theory and optimization, not just in language capabilities.

Startup Hype and Fraud

Brooks draws attention to the pattern of startup hype and fraud in Silicon Valley, emphasizing the importance of maintaining critical judgment and not getting caught up in the excitement of promised "magical" technologies.

Neurosymbolic AI

Brooks predicts that the next big thing in AI may be neurosymbolic AI, which combines neural networks and symbolic AI to create more robust and versatile systems.

Overall, Brooks' predictions offer a balanced and experienced perspective on the future of robotics and AI, cautioning against overhyped claims and emphasizing the need for realistic assessments of the technology's capabilities and limitations.

Rodney Brooks, a renowned MIT roboticist and pioneer, offers a balanced and insightful perspective on the current state and future of AI and robotics. While acknowledging the impressive capabilities of generative AI models like ChatGPT, he cautions against overestimating their abilities and urges a more measured approach to evaluating their potential.

Brooks emphasizes that these AI systems, while highly capable in specific tasks, are fundamentally different from humans and cannot be simply assigned human-like capabilities. He provides examples of how people tend to generalize the performance of these models beyond their actual competence, leading to unrealistic expectations.

Regarding the development of humanoid robots, Brooks draws on his extensive experience to predict that significant breakthroughs are unlikely to occur within the next 25 years. He believes that the form factors and design choices for practical robotic applications, such as warehouse operations, may differ from the common perception of humanoid robots.

Brooks' predictions on the timeline for various AI and robotic milestones, such as dexterous robot hands, home navigation, and elderly assistance, offer a more conservative outlook compared to the often-hyped claims made by some industry figures. He acknowledges the challenges in achieving these advancements and provides specific timeframes for when he expects them to become viable.

Interestingly, Brooks also highlights the importance of maintaining a rational and critical perspective on the startup ecosystem, drawing attention to the pattern of overhyping and even fraudulent behavior that has plagued certain high-profile companies. This broader context serves as a reminder to approach technological advancements with a balanced and well-informed mindset.

Overall, Rodney Brooks' insights provide a valuable counterpoint to the often-exaggerated narratives surrounding the capabilities of AI and robotics. His experience and pragmatic approach offer a more grounded and realistic assessment of the field's current state and future trajectory, encouraging a thoughtful and measured approach to technological progress.