Demis Hassabis, one of the co-founders of DeepMind, said that excessive hype around artificial intelligence is a barrier to making progress in the AI industry.
At the end of last week, Mr. Hassabis, during a conversation with media representatives, expressed the opinion that the high level of popularity of machine intelligence, due to which advanced technology became the object of what can be called global attention, provoked the flood of money into companies operating in the corresponding sphere of activity, and their products, which resembles a public outcry in the area cryptocurrencies.
According to the co-founder of DeepMind, the hype around artificial intelligence is a bit unfortunate. Demis Hassabis says that the high level of popularity of AI in the public space provokes the formation of a situation in which science and research in the context of the development of advanced technology are gradually moving into the category of secondary interest. According to him, in some ways artificial intelligence is not hyped enough, but from another point of view, AI is overly hyped.
Demis Hassabis also stated that in the context of public discourse related to artificial intelligence, attention is paid to unreal things. It is worth clarifying that this does not mean that a kind of general perception of AI as an advanced technology is focused on fantastic scenarios of its use. The implication is that in some cases artificial intelligence is seen as the driving force behind what is impossible. It is also worth noting that the definition of the impossible is in some sense not stable and largely depends on the level of human development in a particular period of the history of its existence. What seemed like a fantastic scenario 200 years ago belongs to the category of everyday life in the space of modernity, the fact of which is not something surprising.
The debut of an OpenAI startup chatbot called ChatGPT has become a kind of catalyst for huge investments in the artificial intelligence industry. The mentioned digital product has become, in a sense, a demonstration of the impressive abilities of machine intelligence. According to CB Insights, investors pumped $42.5 billion into 2,500 startups operating in the artificial intelligence industry last year.
Public market investors have provided financing to the so-called Magnificent Seven technology firms, including, among others, Microsoft, Alphabet, and Nvidia. These companies are leading a revolution in the artificial intelligence industry. The growth of the mentioned technology giants has helped global stock markets achieve such indicators in the first quarter of 2024, which are the highest in the last five years.
Currently, regulators are becoming more active in many countries of the world as part of efforts to monitor the activities of AI firms. In this case, one of the reasons for the increased interest is the issue of security. The significant capabilities of artificial intelligence actually mean the ability to impact many processes in the space of public life, in the sphere of politics, and in the area of economics. Regulators are interested in having a control system that ensures that AI does not become a tool for causing harm, or a source of disasters, and will not be used to develop weapons of mass destruction.
The authorities also demonstrate a desire to monitor companies that provide false or partially unreliable information about their artificial intelligence-based products. Last year, in the United States, the Securities and Exchange Commission (SEC) warned AI firms against inflating environmental records. In March, the regulator accused two investment advisers of distorting their capabilities in the artificial intelligence industry. SEC Chairman Gary Gensler said last month that the emergence of new technologies could provoke a stir among investors and cause false statements from those who allegedly use advanced developments. In this context, it was noted that investment advisers should not deceive the public about applying artificial intelligence models when in fact this is not the case.
The United States Department of the Treasury in March released a study on the cybersecurity risks associated with scaling up the use of AI in the financial services sector. The Treasury has documented a lack of data sharing on fraud prevention. Limited information prevents the development of effective means of protection against cybercrime using artificial intelligence. At the same time, large financial institutions are training AI models based on huge amounts of data.
Demis Hassabis, despite various concerns, believes that artificial intelligence is one of humankind’s most transformative creations. In his opinion, the development of AI is only approaching what will become possible over the next decade. He admitted that, perhaps, at present, in the context of the evolution of artificial intelligence, a new era of scientific discoveries, a new Renaissance, is beginning.
Demis Hassabis says that one of the proofs of the significant scientific potential of AI is the DeepMind AlphaFold model, which debuted in 2021. This digital product predicted the structure of 200 million proteins. Currently, the mentioned artificial intelligence model is used by more than 1 million biologists worldwide.
DeepMind also applies AI to research in the sphere of drug discovery and delivery, materials science, weather prediction, mathematics, and nuclear fusion technology. Demis Hassabis says that his goal has always been to use artificial intelligence as a tool for science.
DeepMind was founded in London in 2010. The company’s goal is to develop artificial general intelligence that would correspond to all human cognitive capabilities. Among researchers, there is an opinion that the mentioned technology will be able to debut in a few decades and there is no guarantee that this is possible in principle. Mr. Hassabis says that one or two more breakthroughs are needed for the emergence of AGI. According to him, the debut of a new type of artificial intelligence is possible within the next decade, noting that the probability of materialization of this assumption is about 50%.
As we have reported earlier, German Companies Invest Over $500 Million In AI Startup Aleph Alpha.
