Diversity Champion

Diversity is something that has made the human race what it is. It has allowed us to solve great problems, survive large apocalyptic events and evolve into what we are today. 

It's something we have labored to entrench, in the greatest sense, into our laws and culture. 

I'm not just writing here on sex, race or other human qualities, I'm talking about commerce, invention, art, science etc. etc.

For example there exists monopoly laws in just about every capitalist nation, simple to protect us, the humans, from a single corporation ruling our lives. The ex-communist states, and how they collapsed, plus the way the whole banking system was brought down, was because that diversity failed or didn't exist.

It has always concerned me how in IT terms, diversity is low. It maybe because in relatively short time we have all progressed off the back of the same ideas, or combined those ideas through large open source initiatives.

None more so than AI, which compared to other initiatives, will absolutely transform society.

I want that transforming to be for the better, not the worse.

So here's my first concern. If everyone basically uses, for want of a better phrase (and to keep this simple), the same methods and code, AI will not be diverse. And if we rely on it, we have created an entity which could be brought down by one event, all of it, in one foul swoop.

The reason plagues fail, and we can use plagues as a mass cyber-attack metaphor, is that genetic mutation and diversity in us all, which means a percentage of humanity survives.  

It's interesting, and as an aside, that recent article around Western culture show that we actually have been recently procreating in a smaller genetic pool. Within very well defined communities, and this limits our diversity. However, that same article pointed out that matchmaking web sites are changing that by redefining communities is a much broader sense so we meet more genetically diverse people! So there you go, maybe AI will help on that score!

Anyways… My second concern is horsepower. That is that the organizations that throw the most compute power at their AI will dominate the market. It’s a bit like Motor Racing where the rules are limited, i.e. cars have a basic set of rules, but the teams can basically throw money at the problem. Just like in Formula One, a team tends to dominate if it has the best design, which usually also includes the horsepower.

Google, Amazon, IBM and Microsoft have that horsepower. Yep that’s four, but will it still be four and is four enough for Diversity?  

And with the race to Quantum Computing where every possible outcome could be calculated by the AI with the most horsepower in the quickest time, what market, or even cultural dominance would that make?

And finally, can we allow AI to make "all the decisions". Should we allow AI to become "emotional" or stay "logical". It's actually the toughest question.

Responsibility, Spontaneity and Creativity are very human things.

So I am a fan of automation… let's get this clear. And I'm not saying AI doesn't have its place, but we do need a debate, quickly on what its place is.

I've heard of people being carried out on stretchers in some Security Operations Centers because the sheer volume of work, attacks and events during a mass attack, overwhelms the analysts and the shift turns into two, three, four… Automating response works very well to protect the humans.

We also know that because of Open Source AI, the hackers have it. They will turn unprotected device, light bulbs upwards, into AI bots. AI will attack. We can watch, or "get more guns" or… we can make sure AI is subject to the same laws that govern people.

And therein lies the crux of this. We need to make sure our Artificial Intelligence is subject to the same cultural, legal and moral disciplines as we are.  

Currently it is not.

Stay safe!

Views expressed in this blog are those of the author and do not necessarily represent the views of any employer or professional body that Mark C Stafford is employed by or a member of.