A computer made from DNA that can solve basic chess and sudoku puzzles could one day, if scaled up, save vast amounts of energy over traditional computers when it comes to tasks like training artificial intelligence models. DNA devices have a number of potential advantages, such as being able to safely store vast amounts of information, in microscopically tiny volumes, for millennia.

Elon Musk's technological predictions may sometimes seem ripped straight from the pages of science fiction. And now the billionaire has made one of his most stunning claims yet, as he says his Neuralink brain chips will one day let paralyzed people walk again. His claims come as Neuralink shared a video of one of the first human patients playing chess telepathically. The chip currently enables the brain to connect with computers, but Musk claims it could one day let the brain communicate with paralyzed parts of the body. In a post on X, formerly Twitter, Musk wrote: 'Long term, it is possible to shunt the signals from the brain motor cortex past the damaged part of the spine to enable people to walk again and use their arms normally.'

Chess has long been regarded as one of the most intellectually challenging games in the world. It requires a deep understanding of strategy and the ability to anticipate and react to an opponent's moves. For years, chess has been dominated by human players who have honed their skills through years of practice and experience. However, in recent years, artificial intelligence has emerged as a formidable opponent on the chessboard. In 2017, Google's artificial intelligence company DeepMind introduced AlphaZero, an AI system that could teach itself how to play chess, shogi, and Go.

Way back in 1985, a team of researchers at Carnegie Mellon University developed a computer purely to play games of chess. After moving to IBM, the computer was further developed, culminating in the obvious test – a match against then-world champion Garry Kasparov. However, the computer known as Deep Blue at this point wasn't enough for Kasparov; it lost four games to two. But like any good underdog, the computer was down but not out. It came back a year later to beat Kasparov in a narrow victory, winning by a single game.

Artificial general intelligence (AGI) may herald our extinction, according to AI safety research. Yet claims regarding AGI must rely upon mathematical formalisms -- theoretical agents we may analyse or attempt to build. AIXI appears to be the only such formalism supported by proof that its behaviour is optimal, a consequence of its use of compression as a proxy for intelligence. Unfortunately, AIXI is incomputable and claims regarding its behaviour highly subjective. We argue that this is because AIXI formalises cognition as taking place in isolation from the environment in which goals are pursued (Cartesian dualism). We propose an alternative, supported by proof and experiment, which overcomes these problems. Integrating research from cognitive science with AI, we formalise an enactive model of learning and reasoning to address the problem of subjectivity. This allows us to formulate a different proxy for intelligence, called weakness, which addresses the problem of incomputability. We prove optimal behaviour is attained when weakness is maximised. This proof is supplemented by experimental results comparing weakness and description length (the closest analogue to compression possible without reintroducing subjectivity). Weakness outperforms description length, suggesting it is a better proxy. Furthermore we show that, if cognition is enactive, then minimisation of description length is neither necessary nor sufficient to attain optimal performance, undermining the notion that compression is closely related to intelligence. However, there remain open questions regarding the implementation of scale-able AGI. In the short term, these results may be best utilised to improve the performance of existing systems. For example, our results explain why Deepmind's Apperception Engine is able to generalise effectively, and how to replicate that performance by maximising weakness.

When we are talking about consciousness, we must understand one thing. Almost all animals in the world have consciousness. That means things like leopards are defending themselves and their descendants if they are under attack. Those animals hunt for eating, but when we are looking at that thing more carefully, we can notice that those cats are making all things in the same way. Reflexes are the things that are automatized actions.

GPT-3 is a 175 billion parameter AI language model that has been trained on a large amount of data. In simple terms, a language model is an AI model that can predict the next set of words given a collection of input words (very much like the auto-complete feature in search engines). Large language models, such as GPT-3, take this a step further by being able to generate source codes or stories based just on a description or suggestion. The startup behind GPT-3, OpenAI, has made its model available to developers via an API. You may sign up for it here, and you'll get a credit of $18.

The pharmaceuticals firm GSK has struck a five-year partnership with King's College London to use artificial intelligence to develop personalised treatments for cancer by investigating the role played by genetics in the disease. The tie-up, which involves 10 of the drug maker's artificial intelligence experts working with 10 oncology specialists from King's across their labs, will use computing to "play chess with cancer", working out why only a fifth of patients respond well to immuno-oncology treatments. Dr Kim Branson, the global head of artificial intelligence and machine learning at GSK, said only 20% of patients respond well to the new oncology drugs that harness the body's immune system to fight cancer. "Sometimes it works like a game buster … and it wipes out the cancer. We'd like that to work all the time. This could be transformative," Branson said.

Machine learning is the science of teaching a computer to solve problems by example rather than writing sequential algorithms which instructions run one by one. Data preparation for machine learning is the prior step towards training a model, and usually involves two substeps: creating a dataset and transforming the data. In this post I'll be focusing on the former in the context of building a human-like AI to play chess in PHP. Because contrary to popular belief, Python is not the only programming language for data science in this world. I am preparing the data on this GitHub repo with MySQL, PHP and Rubix ML, a machine learning and deep learning library for the PHP language.

Melanie Mitchell, a professor of complexity at the Santa Fe Institute and a professor of computer science at Portland State University, acknowledges the powerful accomplishments of "black box" deep learning neural networks. But she also thinks that artificial intelligence research would benefit most from getting back to its roots and exchanging more ideas with research into cognition in living brains. This week, she speaks with host Steven Strogatz about the challenges of building a general intelligence, why we should think about the road rage of self-driving cars, and why AIs might need good parents. Listen on Apple Podcasts, Spotify, Android, TuneIn, Stitcher, Google Podcasts, or your favorite podcasting app, or you can stream it from Quanta. Melanie Mitchell: You know, you give it a new face, say, and it gives you an answer: "Oh, this is Melanie." And you say, "Why did you think that?" "Well, because of these billions of numbers that I just computed." Steve Strogatz [narration]: From Quanta Magazine, this is The Joy of x. Mitchell: And I'm like, "Well, I can't under-- Can you say more?" And they were like, "No, we can't say more." Steve Strogatz: Isn't that unnerving, that it's this great virtuoso at these narrow tasks, but it has no ability to explain itself? Strogatz: Melanie Mitchell is a computer scientist who is particularly interested in artificial intelligence. Her take on the subject, though, is quite a bit different from a lot of her colleagues' nowadays. She actually thinks that the subject may be adrift and asking the wrong questions. And in particular, she thinks that it would be better if artificial intelligence could get back to its roots in making stronger ties with fields like cognitive science and psychology, because these artificially intelligent computers, while they're smart, they are smart in a way that is so different from human intelligence. Melanie's been intrigued by these questions for really quite a long time, but her journey got started in earnest when she stumbled across a really big and really important book that was published in 1979.