Is AI ethical?

In the 21st century we have the privilege of living in a world that is adapting and changing ever faster before our eyes. One of my favourite charts that captures this is the adoption curve from Comin and Hobijin. It visualises just how much faster uptake has become in recent history — what once would take 60 years to reach 80% of the population can now take less than one — and the pace is getting ever faster.

With a pace of change so rapid, it’s rare to get an opportunity and reflect. To ask whether we should adopt a change, rather than just asking if it’s possible.

At Reinventure, we have the privilege of working with some exceptional portfolio companies innovating at the edge of what’s possible. In particular, we have a number working directly in the world of artificial intelligence; Codelingo, Indebted, Basiq, Hyper Anna, Akin and Curious Thing all have AI as their core capability. We are excited about each of these businesses and have a strong appreciation and respect for what the founders and their teams are developing in the space. In particular, we love how each of these companies is looking to use AI for positive societal impact in different ways.

However — as a society, our use and adoption of AI has developed at an alarming rate in multiple fields. Whilst most applications are benevolent, it is worth pausing to consider some of the detrimental effects. This scene from Jurassic Park stuck out from a conversation with a good friend due to its parallels to what we are seeing with the rise of AI:

If we replace “genetic power” with “artificial intelligence”, we see some very significant ethical concerns. In particular:

· Cost — “everyone in the world has the right to enjoy”. We already live in a society that offers vast advantages for those with better access to financial support. Access to private education, private medicine, organic food all come with a price tag too onerous for people in lower income households to benefit from. With AI accelerating growth, the potential for this gap to widen needs to be considered, and ways of making AI decisioning accessible to the public as a whole should be considered.

· Misuse — “it didn’t require any discipline to attain it”. With learning to code becoming ever easier, the ability to create intelligence is ever more accessible. Whilst many use it with ethics and longevity in mind — others may not. Hackers are just as able to leverage AI technology as your bank.

· Consciousness implications — “they’ll defend themselves violently when necessary”. Creating a truly independent consciousness raises many questions. Will it be always aligned to humans? Would it develop rights — just as many of Black Mirror’s dystopian episodes highlight? The lines become very grey. And as perhaps presciently stated in Jurassic Park, “how could we possibly have the slightest idea of what to expect?”

· Job displacement — although not highlighted in this clip, it would be remiss not to mention the scope of potential job losses. In the US alone according to 2014 census data there were 4.4 million Americans working as drivers — that’s nearly 3% of American jobs potentially displaced by the automation of vehicles alone. Whilst AI will inevitably also create new jobs, society needs to consider how to bridge skill gaps and avoid entirely displacing such large masses of people.

“so preoccupied with whether or not they could, that they forgot to think if they should”

Whilst I’m not making a case against AI — on the contrary I think it can be used to significantly improve our lives and improve society’s standards of living — it is worth considering that the pace of technological change now far outpaces our laws (consider all the incoming regulation in the blockchain space), so it must be undertaken with a clear consideration of ethics and the broader applications and impacts to our society. Playing with such powerful technology that broadens human possibilities should be done with consideration for second and third-order effects — not just the immediate benefits to a primary use-case.

Onwards and upwards — but mindfully.