On October 30th, Chen Tianqiao announced at the first AI-Driven Science Seminar of Tianqiao Brain Science Research Institute that he would invest a new $1 billion to support innovative artificial intelligence research of scientists around the world. It sounds a little mysterious, but it smells like a master of strategy in his bones, and it is insanely practical. Why do you say that? We have to break some boundaries to see. Original: Chen Tianqiao announced a $1 billion investment to fully support the research of “discovery intelligence”.
In the past, when we talked about artificial intelligence, we always felt that it was a super scholar. If we gave it a hundred questions, it could answer all the mistakes correctly. That was called computational intelligence, or task intelligence. It was good at solving problems that we already knew the answer or the path to solve, such as writing programs, translating, playing chess, and leveraging efficiency.
But the AI that Chen Tianqiao wants this time is not to do the problem, but to “realize the whimsy”. In his words, the ultimate value of artificial intelligence lies in “discovery”, which is like a scientist who is bursting with curiosity and especially loves to be serious. He will take the initiative to build a theoretical model about the world, then he will put forward a hypothesis himself, and then go to various scenes to see if his hypothesis can be beaten. It quickly corrects its worldview … This is an infinite cycle and self-iterative process, which is the secret path to general artificial intelligence (AGI). Think about it, a top scientist. What’s so good about him? It’s not that his memory is better than others, nor that his calculation speed is faster than others, but that he can see at a glance the wrong place, the blind spot that can be asked and falsified, in a lot of data and phenomena. Newton saw the apple fall to the ground, and he asked: Why won’t the apple fly to the sky? Instead of: this apple is so sweet! Chen Tianqiao’s ten billion dollars is the ability of AI to learn to ask this “100,000 whys”. This kind of discovery-type intelligence needs particularly intriguing abilities, which can not be solved by simple calculation. It needs a neurodynamic structure to support rapid learning and adaptation, and a long-term memory ability to remember all the somersaults it has fallen. Crucially, it needs a causal reasoning mechanism and a complete world model, and it needs to know “why butterflies flap their wings and cause it to rain in Texas” instead of simply knowing “it rained in Texas after butterflies flap their wings”. Finally, it needs a metacognition and internal motivation system. To put it bluntly, it needs to know what it is thinking and have inexhaustible curiosity to explore the unknown. Isn’t this the core competence that students generally lack in exam-oriented education? ) This is simply to disassemble the thinking structure of a top scientist into several clear parts, and then incubate them with billions of dollars of resources. This business is simply to turn the thinking process of high IQ into a quantifiable project! Some people may say that a billion dollars sounds a lot, but compared with the technology giants who burn graphics cards for tens of billions, it is a drop in the bucket. But this is the genius of the strategist. Mr. Chen’s money is not sprinkled on the simple and rude computing power, but on the two leverage points of “mechanism” and “talent”.
Mechanism First of all, he announced that he would establish a cross-disciplinary benchmark evaluation suite, the core indicator of which is “discoverability”, which is equivalent to setting a new rule for AI researchers all over the world. In the future, it is not a skill to answer the questions correctly, but depends on who can put forward more beautiful, innovative and subversive assumptions. This has set the research direction at once, which is more ingenious. It is much more cost-effective to use the mechanism to incite global wisdom than to hire people by yourself. The second is talent incubation. PI incubator is set up to directly let doctoral students and postdoctoral students have independent portals, give them resources to set up their own laboratories, and pursue innovative ideas that sound “irrelevant”. You know, many great discoveries in the history of science started with “whimsy” at first, but young people often lack start-up funds and an independent stage. Chen Tianqiao’s move directly targets the pain points of young people’s scientific research and provides them with a high degree of freedom for trial and error. Think about it, isn’t it because of this “discovery intelligence” that many things around us are successful? Just like in the old street of my hometown in Wuzhou, there is an old woman who has been selling rice flour for decades. The reason why her rice flour tastes better than others is not because she uses any high-tech machines, but because she tries a new soaking time of rice flour and a new fermentation temperature of sour bamboo shoots every morning with anticipation, and even puts an extra star anise today and a few drops of oil tomorrow … She silently constructs a world model of “the most delicious rice flour” and keeps on. Chen Tianqiao’s investment is to hope that the future machines can also have the wisdom of striving for perfection, constant hypothesis and verification, but at a much faster speed and on a much larger scale. Therefore, the significance of this billion dollars has already surpassed the money itself. It represents our profound shift to the essence of intelligence, from the pursuit of “faster” calculation to the pursuit of “deeper” cognition, from relying on AI to solve known problems, to let AI lead us to explore the unknown. The future scientific research is no longer a single-handedly by scientists, but a new chapter in the cooperative evolution of human intelligence and discovery intelligence. Chen Tianqiao, a former game tycoon, has now played the most ingenious strategic layout in the fields of brain science and AI. His investment is not technology, but human curiosity. What he pursues is not the completion of the task, but a breakthrough in wisdom. This highly leveraged investment method with global talents and new rules is brilliant! The “discovery intelligence” mentioned above is also the design concept of “LLL Insight Agent” that our LLL Ai studio has devoted itself to creating. We clearly know that the AI in the future will not replace human beings, but become the strongest advisor for human beings to explore the unknown world-when a vague and marginal new discovery flashes in our customers’ minds, they can directly input the fragment of this “discovery” to the twilight insight agent.
