AI will not take over the world
In a recent blog we discussed the reliability of AI and automation and the fact that these systems are devised by human beings, highly skilled human beings to be sure, but human beings make mistakes.
Wired has just published two further articles exploring the issue of AI.
In the first it explores whether it is possible to make AI technology completely unbiased and also asks how many businesses benefit as much as they could from AI technology.
It reports that the return on business investment in AI has declined by 27 per cent over the last five years.
The reason, it argues, is that “companies don’t know how to make the most of AI and data analytics, and how they can apply to business problems.”
It also suggests that businesses get things the wrong way wound when considering investing in AI, so that they under-use its potential. It advises that businesses should “start by drawing up a list of business challenges and prioritise them by whether or not they can be addressed by using AI and the expected return on investment”.
The second article, by Joi Ito, director or MIT’s Media Lab, questions the assumption that AI can and will supercede humans in almost every sphere of activity.
Ito calls this assumption singularity in which those people who have succeeded in mastering the power of AI capture all the wealth and power.
This, Ito argues is “reductionist” thinking and only works for a very narrow range of learning and thinking which can lead to over-simplified ways of “fixing” humanity’s problems.
However, Ito says, most of the challenges we face today, such as climate change, poverty, chronic disease or modern terrorism have actually been the result of this reductionist thinking and we need to respect that many human problems are actually much more complex.
Machines, and therefore AI, need to be adaptive and to augment, not replace, humans. “not artificial intelligence but extended intelligence”.