Quantum Processing: The Path from Autistic to Empathic AI?

When the challenge of building an artificial intelligence has been called everything from "dangerous" to "the last invention man[kind] ever need make," it's natural we'd want to look at the problem from every possible angle. In previous articles, we’ve explored the art and science of building better machine intelligence, and at how male and female partnerships have incredible potential for advancing our technical progress.

Now, taking a step back from the milestones and processes checklists associated with both company building or AI research, let's look at the high level picture of building a more human, more empathetic AI that can solve really hard problems - that is, challenges that are simultaneously difficult for humans and machines at present, like optimizing America's healthcare system.

State of the Industry

As we talk about the intelligence of machines and of people, it helps to keep a perspective on where we currently are in our empathetic AI design -- which according to IBM Watson's creator David Ferrucci, is not very far:

(Watson) is akin to a human autistic savant.

Pulling emotionality into our technology shouldn't be so hard, right? It's said that Silicon Valley is supposed to full of makers crafting products that are both useful and delightful to users across the world. What it shouldn't be, then, is Moneyball author Michael Lewis calls,

A bunch of autistic people wandering around. It's a hard place to write about because there's a lack of emotional content. It's a cold place.

So what is causing the greatest barriers to innovating on less autistic, more authentic systems? What questions do we need to be asking our teams and investors as we pursue work not just in artificial intelligence, but in other important technologies? And last, how do we get the best minds involved?

Scientific Reasoning: First Understand, Then Advance

When we read the ocean of articles on AI -- such as Vicarious’ attempts to build a single unified system that will be generally intelligent like a human -- it may not be immediately obvious that developments in quantum physics and neuroscience can inform and affect how we go about building such systems and which functional components should be factored in as part of intelligence.

In layman's terms, think of building an artificial intelligence like the mission to land on the moon: it took the combined efforts of entire nations in countless domains - from material science to nutrition to chemical engineering to sanitation - to make the journey happen. In AI work, the four biggest components are two distinct types of hardware and software:

The first is technical hardware for machine processing - that is, microprocessing, quantum processors, and more.

The second is technical software for simulating the mind, ideally taking into account both the rational and empathetic aspects of human behavior entirely generated through 

The third is biological hardware - especially the brain and body - that are the inspiration for our existing models of intelligence, which we have yet to replicate in a meaningful way.

The last is biological software - namely, the mind - which is still poorly understood yet which teams of researchers are attempting to replicate.

Each aspect demands being understood. While the developers amongst us tend to default to the immediate pragma of “What programming language can we use?”, “Which functions, variables and classes need to be applied?” and “How can we build a graph from the data sets?” whereas we should ask the First Principles questions:

The First Principle Questions of Creating an Intelligence

  • What is intelligence?

  • How do humans learn and become intelligent?

  • What are the factors that make us intelligent and different?

  • When did our intelligence undergo the most transformation --- was it during the invention of flint tools, the Industrial Revolution, Moore’s Law etc.?

  • Was our intelligence triggered by social needs? Language? Emotions? Documenting our senses? Daily lives? Transactional behaviors?

  • When babies learn, is it in linear and sequential isolation? In parallel? Via simultaneous stimuli?

  • Where in our minds, bodies and souls is intelligence stored, filtered and adapted?

  • Why does intelligence even matter? Is intelligence mere atomic matter or is it more?

  • Does intelligence also have emotions, language, culture and consciousness in its very DNA?

That humans are more intelligent than the ability to play chess and to pass IQ tests is intuitively obvious but not yet modeled or measured coherently enough in the philosophical and scientific frameworks we use to define intelligence.

A lot of those philosophical and scientific frameworks originated in the 17th and 18th Century works of Rene Descartes, Blaise Pascal, Pierre de Fermat and Thomas Bayes. It’s their models for linear rational thinking that has underpinned a lot of our mathematical models for intelligence, including Turing’s ‘Can Machines Think’ and the Stanford-Binet IQ tests, for hundreds of years.

Borrowing from Aristotle’s logic, Pascal and de Fermat proposed that there can only be one of two discrete binary outcomes:

  • 0 = an event happens; and

  • 1 = an event doesn’t happen

with a probability likelihood between 0 or 1 of any deterministic event occurring. 

Summarily, our models of intelligence remain poorly understood because they reflect only the rational side of thinking.

Notably, Einstein’s Theory of Relativity of 1905 and Schrodinger’s Cat paradox of 1935 started to challenge the ideas of binary (0 or 1) events and whether the Universe and its information is only probabilistic. In a 1943 conversation with William Hermanns recorded in Hermanns' book Einstein and the Poet, Einstein remarked:

As I have said so many times, God doesn't play dice with the world.

If the basic premise of probability is awry according to Einstein’s theory, what does this mean for all the applications we’ve relied on probability to solve from modeling people’s behavior and their employability to their social graphs and knowledge graphs?

The Beginnings of “Bye bye, binary?”

This year, over a hundred years on from Einstein's Theory of Relativity paper being published and sixty years on from John Von Neumann and John McCarthy's models for classical computers, research from the University of Southern California and D-Wave, the Canadian makers of quantum servers that supply Google and NASA, caused Wired to explain this:

A quantum computer operates according to the principles of quantum mechanics, the physics of very small things, such as electrons and photons. In a classical computer, a transistor stores a single “bit” of information. If the transistor is “on,” it holds a 1, and if it’s “off,” it holds a 0. But in quantum computer, thanks to what’s called the superposition principle, information is held in a quantum system that can exist in two states at the same time. This “qubit” can store a 0 and 1 simultaneously.

Two qubits, then, can hold four values at any given time (00, 01, 10, and 11). And as you keep increasing the number of qubits, you exponentially increase the power of the system.

Another area of Quantum Physics worth following because it speaks to how subjectivity and intuition are as necessary as objectivity and rationality to how the Universe, and perhaps our minds, work are contained in the physicist, Max Tegmark of MIT’s, ideas on ‘Consciousness as a state of matter’, in particular:

Perceptronium, defined as the most general substance that feels subjectively self-aware.


As Michael Lewis has argued, Silicon Valley is “autistic”; this means the products and systems being produced there have almost no constructs of emotional subjectivity, which may have been exacerbated because of echo chamber effects in Silicon Valley.

Lady Martha Lane-Fox, the co-founder of Lastminute.com, noted earlier this week:

All that’s happened is that one bunch of very rich white men have transferred their money to another bunch of very rich white men and, worse than that, they are in a very small concentrated area of the world, in Silicon Valley.

There is a cycle of behaviour in the venture capital community which I don’t think is overt sexism, I think there is some, but I don’t think it is the only reason but there is a lot of unconscious bias.

If you are a venture capitalist and you are looking at risk you are less likely to invest in someone that is not like you.

There are a myriad of articles on how male investors don’t invest in female founders and how the low 4% of VCs at the top firms being female result in missed opportunities for investors and founders, alike.

So as we step-up to make systems better, the most important question for all of us is how to value the “differences that make a difference” and get us towards an integrated model wherein system specifications are made by men and women, working well in synch.

Furthermore, we need to think through how this would be transformative for the next wave of technology which is increasingly seen to be around AI. It’s time to pool together to make the next leaps in innovation a reality. Quantum Intelligence is within our minds and hands to create so let's get busy and build those trillion dollar quantum ideas.