Should We Really Be Concerned About AI?
There is a growing trend of scaremongering about the dangers of AI. Ill-informed opinions are bandied about singularities and emergent abilities etc. Journalists and content creators are engaged in a feeding frenzy; prophesying doom and suggesting more and more radical and solutions.
With all the commotion currently surrounding Artificial Intelligence it can be difficult to get a clear picture of where we really are and what to expect in the next few years. It seems that every day a new AI innovation is reported that will change the course of human civilization for good or for ill. As always, the reality is a little more complex and we have to pick through the marketing hype, media spin and content creators chasing views through alarmism.
Predicting Horse Races using Artificial Intelligence
AI has proven particularly valuable in handling vast sets of noisy data. By noisy we refer to data that contains random inconsistencies such as in horse races. Some races will always end in unpredictable results – an out-and-out favourite may fall at a fence handing a long odds horse an unexpected victory. Such results can be problematic because it can obscure true patterns in the data. If this were not the case then AI would not be required; a formula could be found that would mean the end of the gambling industry in sports.
There have been many implementations of applying AI to sporting results such as https://racingoracle.com. Since 2016 they have been using AI to predict the results of horse races. The approach they take is to create statistics from a large dataset of races; with variables such as percentage wins, places and days since last race. By leveraging AI in this way, they aim to make more informed decisions when backing winners and developing race strategies. The bulk of the research is in the gathering and preparation the of data.
What AI Isn’t
It’s understandable to assume AI is a single, uniform technology, but in reality, Artificial Intelligence encompasses a diverse range of technologies at different stages of advancement and with varying utility. For example, research into computer vision is quite separate from natural language processing and different again from self-driving cars.
Part of the panic around AI has been the apparent rate of innovation. In fact, machine learning and AI have been around for decades. We are now seeing the fruits of many years of research rather than overnight revolution. A case in point is the plethora of YouTube videos about self-driving cars. Content creators have discovered a world changing brand new technology that only a brief study would show have been in trial for over 50 years! Artificial Neural Networks, the fundamental technology behind both Tesla self-driving and Google’s Waymo, came from research that started to find its feet in the 1950s.
It is possible the reason for the current focus has been brought about by advancements in AI content creation. This is the area that has strayed into the territory of the media – both traditional and digital. ChatGPT and other large language models can produce impressive content in seconds. Large language models again use artificial neural networks and comes from research into Natural Language Processing itself many decades old. Deep fake videos and images can be difficult to discern from the real thing.
Future of AI
The AI above tools should not be confused with Artificial General Intelligence. This is the real area to be concerned about – think Terminator and HAL 9000. Thankfully this does not look close to fruition and may fall into realm of technologies that humanity seems to edge closer towards but never fully achieves. Fusion Power and Quantum Computing may fall into this camp too.
In any case AI certainly provides great opportunities. It may be the case that consumers of content are more concerned with the accuracy and standard of the content than the source. With the lowering of the bar for today’s media it may be seen that a revolution in content creation will result in a better informed public at the cost of more humanities graduates working in the service industry rather than leading public debate.
