Mainstream media doesn’t get AI

Source: Deep Learning on Medium

Tech journalism needs a reboot

Photo by rawpixel on Unsplash

A lot of technology journalism is driven by hype and exaggeration at a time when we desperately need context, perspective and education to get to grips with how the world is changing. Nowhere is this more evident than in the coverage of artificial intelligence (AI).

A recent example can be found in the reports about a fantastically brilliant chess-playing programme called AlphaZero, developed by the Alphabet-owned (Google’s parent) AI research company DeepMind. It is important to stress that the news outlet linked to above is neither worse nor better than countless others and only selected here because it was the top article in the search engine results.

It is very likely that the reports are all based on the same corporate PR material. They certainly all claim that AlphaZero has shown signs of “human-like intuition and creativity” in what is invariably described as a watershed moment in history.

The story is that AlphaZero has developed a new style of playing chess which is much closer to human improvisation than traditional computer chess. That is because AlphaZero learns from its past successes and mistakes, rather than calculating millions of possible permutations as it plays. For the record, according to Wikipedia, AlphaZero searches 80,000 positions per second in chess, compared to 70 million for the Stockfish chess engine.

AlphaZero uses (deep) neural network technology, made possible over the past decade by notable improvements in machine learning, sometimes called deep learning. As computing power has increased, deep neural networks have made machines capable of performing tasks in a way that would not have been possible using traditional programming techniques.

This has transformed technologies such as computer vision and natural language processing (NLP), which are nowadays being deployed at a massive scale in many different products and services. Manufacturing, healthcare and finance are just some of the sectors that use deep learning to uncover new patterns, make predictions and to guide decision making.

It is a fantastic piece of technology, but there is nothing human-like about AlphaZero’s chess playing exploits. The programme is uniquely focused on the task of playing chess and has only the ability to recognize patterns in moves and positions and to act accordingly.

Developers differentiate between ‘strong AI’ and ‘weak AI’. Strong AI, sometimes called general AI, refers to a machine able to solve any problem requiring advanced cognitive abilities. It would be able to deal with new situations and solve problems it has never faced before, not confined only to chess.

That is the technical explanation. Another way of looking at it is that unlike humans, AlphaZero has no consciousness of its actions. In other words, AlphaZero is not only unaware that it is playing chess, but also unaware about anything else. It is capable only of performing pre-programmed tasks, however brilliantly.

Earlier this year, an article in the Columbia Journalism Review called for a new model for tech journalism. It ends with the advice that we should “let companies do their own PR, even if this comes at the cost of clicks.”