For all others: I
will here share my experience and thoughts from the three days in Geneva during
the first global gathering on the topic of #AIforGood.
Supported by the
UN, UNESCO and other NGOs, ITU hosted an impressive conference in Geneva in
order to discuss the ethical, technical, societal and policy issues related to
AI. Furthermore the goal was to come up with guidelines and recommendations how
to proceed with AI innovations.
As one of the 600
invited guests, my quest and motivation was:
- first to understand
the global status quo on the topic
- second to nurture
my initiative that design is the single most important steward in developing
the ethics of AI
Marcus Shingles of
Xprize kicked off the first day with infusing the audience about the
unimaginable impact of exponential technology growth/development. Not new but I
realized over the days that many participants were neither innovation gurus nor
technologist. So his evangelism was quite important. And: exponential effect
remains a crucial factor in AI since for now we are still at the bottom of the
curve and who knows when the explosion kicks in.
So here come my
most important insights:
- We are still very far from human-level AI. For sure it is impressive to see a smartphone saying aloud all the objects it gets presented in front of its lens in staccato. And yes, awesome the accomplishments Audi CEO Rupert Stadler described in his talk. But this has nothing to do with real intelligence as we know it from humans. Technology on real strong AI has hardly advanced since the 50s of the last century. All the developments we experience today are the result of the increased processing power of data. Deep learning for example is good at certain aspects of perception, particularly categorization (important for text or image recognition) but perception is more categorization and cognition (intelligence) is more than just perception according to Gary Marcus. Sometimes AI is also just a show. This became evident when David Henson presented Sophia. A remote controlled puppet giving us the illusion that humanoid robots are just around the corner.
- People don’t know exactly what AI is. For some it is the utilization of big data and for others its killing drones. If we want to build up rules and guidelines we have to make sure, that we have a solid definition of the subject. Unfortunately, the conference fell short on this.
- Governments and NGOs are fully aware of the subject and are acting. I must say I was concerned this might not be the case. Mady Delvaux-Stehres, Member of the European Parliament proved that the EU is taking proactive measures on the ethics, legislative matters and technology funding.
- Mind the social impact of AI. It doesn’t come as surprisewhen hundred representatives from NGOs (mostly UN, WHO) participate, but I just didn’t consider this enough. The social impact through AI is huge, and multidimensional. For one dimension take the socialist scenario often discussed: what if one day we don’t need to work for money anymore because robots do the working. We all get an income and can strive for more important matters. But how would you share the wealth existing today? How would we transition from today’s unfair wealth distribution to a more equal one? And would AI be of any help? Another dimension: who owns AI intellectual properties? Private Companies? Governments? Who owns the date given privately for a bigger cause (e.g. health data)? Will AI increase inequality and discrimination? Because today it is doing so. Tons of questions and little answers. Thanks to Geneva this will be on my agenda too.
- One plus one makes three. Human-machine teams are the future. There was a great and impressive example where AI and human joined forces in the medical field and outperformed by far their individually achieved results. As long as we have no human-level AI we should focus on developing symbiotic solutions, where human power gets excelled through AI- not replaced. This would build up trust in a new and sometimes scary technology world. So watch for Cobots in the future.
- Human-centricity is global. Although nobody really talked about design, everybody was talking about creativity, empathy, collaboration/co-creation with stakeholders and human centricity. Literally in every talk. This design minded approach finally has become a global trend and perfectly fits in the ethical discussion around AI. Katsumi Emura CTO from NEC even stated that Society 5.0 will be about human-centricity (whereas Society 1.0 was about hunting & gathering).
So what will I take
away from this first global conference? What is truly important to me?
- Our human-first approach at INDEED is valid. More then ever we need to empower humans. And AI is an excellent booster to make humans even more powerful and important to live in a better environment and society.
- Despite everybody talking about the importance of creativity, empathy and human-centricity hardly anybody talked about design or design thinking. Obviously our design community has still a long way to go.
- Don’t forget the forgotten ones. I will try to think more about the many that can not speak for themselves.
Please feel free to
drop me a line if you disagree or if you have something to add.