The need to take one step at a time

“Take the first step in faith. You don't have to see the whole staircase, just take the first step.”

― Martin Luther King Jr.

Having covered the basics of how AI and deep learning methods such as CNNs work, it is worth addressing the question of how we can usefully apply such technologies for significantly advancing science and technology in critical areas. The important points here are that we need to take one step at a time and data is king ; or perhaps I should say data are king, since data is a plural (the singular is datum). For example, computers are very good, and have been for some time, at playing games. A famous example is provided by the series of chess matches that were played between the computer known as ‘Deep Blue’ and the chess grandmaster Garry Kimovich Kasparov. Such a six-game chess match was played in New York City in 1997 and won by Deep Blue. This 1997 match was the first defeat of a reigning world chess champion by a computer under tournament conditions. A mistake that is often made is to then assume that the computer concerned actually has human intelligence and can think , so that it would be able to replace a human expert in undertaking some complex reasoning tasks. By the way, I am not saying that computers will never be able to think like humans – if this were possible no doubt some great advances could be made, but we are not yet at such a stage. When discussing whether computers can think, I always tend to be reminded of the Star Trek episode, ‘The Ultimate Computer’. In this rather good episode, talented computer scientist Dr Richard Daystrom presents his new computer - the ‘M-5 Multitronic System’. This is an advanced artificially intelligent control system that is said to be able to control a starship; as Dr Daystrom put it: “The M5 thinks, Captain”.

The Enterprise, with a skeleton crew, is ordered to test how the M5 performs in a series of war games. Kirk is, as you can imagine, not best pleased by the concept of being made redundant by a smart computer. If the M5 was able to think it must have had some serious (artificial) mental issues, since it starts destroying the other starships in the war games, thereby killing many people. Daystrom has a breakdown and it is left to Kirk to talk the M5 down and save the day (which is perhaps a somewhat too commonly employed means for solving issues presented in a number of Star Trek episodes).

What all this might indicate is that getting a computer to think in a truly useful and reliable way may well be quite a long way off – Star Trek was set over 200 years in the future and it seems that even at this stage in man’s progress, genuine and reliable artificial intelligence had not been attained.