Skip to main content

AI Metamorphosis: 4 Mind-Bending Things to Add

By July 18, 2019Article

In the garden over morning coffee, I read, “The Metamorphosis: AI will bring many wonders. It may also destabilize everything from nuclear détente to human friendships. We need to think much harder about how to adapt by Dr. Kissinger, Dr. Huttenlocher and plain ol’ no PhD Eric Schmidt. It’s brilliant. 

Then, I thought of four mind-bending additions to this piece – that I hope are directionally correct but really just thought-provoking rather than claiming to detail the near term future. 

1. Can AI ‘Do’ Philosophy

Massive AI does not use the language of humans, our frame of reference or what we call knowledge – and never what we would call wisdom. It is just not that in any way. Nor does it really emulate our brains which while massively complex and huge (from a node count) are very slow if not useless for the scale ‘problems’ AI loves.

Philosophically, it has no epistemology nor will it ever. That doesn’t just mean it has a hard time really understanding the difference between well-founded belief and opinion it means it doesn’t even ‘think’ in those terms. So we really shouldn’t be so surprised that an AI heavyweight learns to play chess in 24 hours in a way that makes little sense to chess masters and minor detail becomes best in the world in the same day. That is what it does. It blasts through scale complexity in new innovative fast ways that don’t really have human reference points because it’s not human. That makes AI and ethics a really murky area.

2. Digital AI Companions

Old frames of reference won’t be right. The end game isn’t digital companions who are wicked smart or even digital twins which are really useful. Near term, we will see more of both. Except for hermits, the entire world is humans in interactions engaging with things and information. These processes, patterns and progressions will all become more proactive, preventative, pattern matched, personalized, predictive and hopefully in many cases permissioned.

All of this means ‘life’ becomes more intelligent with increasing efficiency, quality, satisfaction, engagement, cycle time, knowledge and hopefully understanding. The great gift is it will eliminate time waste and give us time back…our scarest resource. That is so much more than a digi-companion. Humans and the landscape they move through will simply be augmented by AI in a myriad of ways. AI will deliver Evolutionary Intelligence where we continually improve as the environments we move through harvest data exhaust and apply that to whatever we deem progress to be.

Homo Sapiens will become Homo Cumulatus within one generation. A short piece on that:

We will move beyond just data analysis and clever math, machine learning, predictive analytics, etc. We will watch networks of human and synthetic data behave. It’s that behavior that will create new questions and opportunities. 

3. War, War Fighting and Optimism

 AI makes me more optimistic about war for three reasons:

 1) It will become more precise, predictive, preventative… AI will help leading powers see the digitized landscape and outcomes better and hopefully reduce mistakes and overreactions that have historically been the hallmark of war and indeed the triggers for escalation. It will also be because we will see consequences more clearly. Gone will be the days where an Arch Duke taking a wrong turn triggers anything and especially not the scale of WWI. 

2) WMD’s will be overshadowed by WMI’s (Weapons of Mass Impact). These will be less kinetic and bloody which in itself is a cause for optimism. Unfortunately, they may also be worse. An infrastructure attack clogs ports for 4 months that happens three times in two years. An EMP over Silicon Valley or just enough kinetic incidents at your favorite burger chain kills that business for a decade. All the HVAC systems get killed in a large desert country in summer. Hospitals, police departments, EMS without compute/connectivity even for short periods is a nightmare. 

3) AI will make war fighting become a mosaic game of smart interacting lego-ized modules that are fit for purpose and polymorphic in defense. The utility, compartmentalization and agility will make actual engagement less palatable, less effective. It will make big wars less winnable and small ones really short. Clausewitz said ‘war is the continuation of politics by other means’, hopefully, this extends the continuation of the non-war part of politics. Of course, this will concentrate power into the few that can do this at scale. No disrespect to Moldova (or Marvin Zonis) but their military will not feature in this next scenario. 

4. The Next ‘Worry’

 ‘Everyone should learn to code and to program.’ That was somewhat useful a decade or two ago. AI will turn on itself. AI will turn and focus on the IT/computational world and radically improve the field. In fact, it already is.

Networks and processing will be really hard fields as scale and speed demands accelerate but smart people/companies will manage that. But coding/programming will turn into LEGO assembly and often will just be automatic output from Service Design or process layer work. If only there was a no-code/low-code movement taking off…oh wait, yep we have one of those.

Last year I saw three young women interns from Cornell Tech build an amazing complex solution in 2 weeks using code LEGO. That is a glimpse of the beginning. Infused with ever-accelerating AI, what does a world look like where the code layer is automatic and computational intelligent solutions are almost spontaneous? But you need not worry, that is the distant future, like three years from now. We will need more architects.

My Positive Conclusion

This change will revolutionize education and healthcare and blast through some of the flat morally wrong socio-economic disparities and self-inflicted societal wounds we have to today in those fields. Better educated, better connected, healthier humans is by itself a great thing.

That alone makes this the best of times.

But Eric is right this AI thing needs adult supervision and some analog of the bioethics approach we see emerging. The tricky thing is that would require humans and they are most certainly not designed for that. 

Toby Redshaw is SVP at Enterprise Innovation and 5G Solutions at Verizon

Copy link
Powered by Social Snap