An Ubtech Walker X robot plays Chinese chess at the World Artificial Intelligence Conference in July of 2021.
Artificial intelligence is still in its infancy, despite the fact that machines are getting smarter and smarter every year.
Artificial intelligence can excel at certain tasks, but it can't do more than one thing well. A seven-year-old has far more intelligence than any of today's artificial intelligence systems.
Edward Grefenstette, a research scientist at Meta Artificial Intelligence, told CNBC that artificial intelligence is good at approaching individual tasks.
The real world has significant potential for change, but we are bad at capturing it within our training.
Grefenstette said that there are ways to efficiently adapt training methods to changing environments, resulting in more robust agents. He thinks there will be more industrial and scientific applications of such methods this year.
Artificial intelligence has a long way to go before it can do things like human-level intelligence, but companies like Amazon, Facebook, and Google are still investing billions of dollars into hiring talented researchers who can potentially improve everything from search engines and voice assistants to aspects of the internet.
Beth Singler, anthropologist at the University of Cambridge, told CNBC that claims about the effectiveness and reality of artificial intelligence in places that are now being labeled as the metaverse will become more commonplace in 2022.
There could be too little discussion about the effect of the metaverse on peoplesidentities, communities, and rights in 2022.
Gary Marcus, a scientist who sold an artificial intelligence start-up to the ride-sharing company and is currently the executive chairman of another firm, told CNBC that the most important breakthrough in artificial intelligence will be one that the world doesn't immediately see.
He said that the field of deep learning still has a long way to go. Deep learning is an area of artificial intelligence that tries to mimic the activity in the brain to learn how to recognize patterns in data.
Marcus believes that the most important challenge for artificial intelligence is to combine all the knowledge of science and technology with deep learning. He said that deep learning can't leverage all that knowledge and is stuck again and again trying to learn everything from scratch.
Marcus said that there will be progress on this problem this year that will eventually be transformational, but that it will be another few years before we see major dividends. The first medicine in which artificial intelligence played a significant role in the discovery process is likely to be seen this year or next.
DeepMind, which is owned by Alphabet, is one of the biggest artificial intelligence discoveries of the last couple of years.
The company has created an artificial intelligence software that can accurately predict the structure of a molecule in a matter of days, which could pave the way for better understanding of diseases and drug discovery.
Neil Lawrence, a professor of machine learning at the University of Cambridge, told CNBC that he expects DeepMind to target more big science questions in the future.
Artificial intelligence systems that can generate convincing text, converse with humans, respond to questions, and more are set to improve in the year 2022.
DeepMind said in December that its new language model can beat others 25 times its size.
Catherine Breslin, a machine learning scientist who used to work on Amazon's voice assistant, thinks Big Tech will move toward larger and larger language models next year.
Breslin told CNBC that there will be a move toward models that combine vision, speech and language capability, rather than treat them as separate tasks.
A new breed of companies will likely use language models to predict the most effective ribonucleic acid sequence, according to Nathan Benaich, a venture capitalist with Air Street Capital.
He said that last year the impact of RNA technologies was seen as novel covid vaccines brought an end to nation-wide lockdowns. I think we will see a new crop of companies this year. These new companies could dramatically speed up the time it takes to discover new drugs and vaccines by using language models to predict the most effective RNA sequence to target a disease of interest.
There are major concerns about the ethics of artificial intelligence, which can be highly discriminative and biased when trained on certain datasets. Artificial intelligence systems are being used to make fake porn.
Verena Rieser, a professor of artificial intelligence at Heriot-Watt University in Edinburgh, told CNBC that there will be a stronger focus on ethical questions around artificial intelligence in 2022.
She said that she didn't know whether the artificial intelligence would be able to do much new stuff by the end of 2022, but that it would be more inclusive.
There will be revelations about the use of machine learning models in financial markets, spies, and health care, according to an independent researcher who used to work for a Big Tech firm.
He told CNBC that it would raise major questions about privacy, legality, ethics and economics.
