An experiment.
Hold out your right arm and then slowly flex your wrist.
Which comes first – the thought to flex your wrist, or the action itself?
In the 1980s, an experiment by the neurologist Benjamin Libet raised some disturbing questions about the extent to which we can say we are conscious before a decision like this.
In his lab, Libet hooked a group of people up to an EEG machine, measuring brain activity via electrodes on their scalp, and making use of a precise timer, he asked the volunteers to record the moment they became aware of the urge to move their wrist, discovering an average delay of 200 milliseconds before the action was taken.
But the recordings also revealed another signal – 550 milliseconds earlier – that suggested the brain was actually ready to act long before we are conscious of it.
How much free will do we actually have in making our decisions?
A prediction for 2030
We’ve been thinking about this experiment recently in the context of technology.
The question is: to what extent will we allow artificial intelligence to guide us in our daily decisions?
We are convinced we will soon be able to give machine learning algorithms access to our intimate thoughts and impulses, allowing biosensors to relay useful data to networks where they will be parsed and processed with millisecond delays, presenting us with a branching tree of decisions about almost every routine in your day – what to eat, what to listen to, to whom we speak, the small lessons we might learn to incrementally adapt our lifestyle.
Mark Zuckerberg, for one, speaks of an almost telepathic means of communication by 2030, “we are going to be able to communicate our full sensory experience to someone through thought via head wear than can scan our brains and then transmit our thoughts to our friends much as we share baby pictures on Facebook today.”
How do we get there?
We’d argue that, for professional investors, the key to negotiating the next few years of wild industrial weather is to focus on technologies that produce the most useful and actionable data about our lives.
This idea of focusing on the actionable data is a very useful one for investors. It cuts through everything.
All the technologies that are now emerging — sensors, machine learning algorithms, vast computing power in the Cloud, mixed reality — are at the service of producing actionable data for the big internet dogs.
And we can follow this idea of genuine information as it mutates from: Big Data (largely useless or at least confounding) to Actionable Data (a much smaller set) to Anticipatory Data to Augmented Intelligence (with huge networks of autonomous machines hitching together from different domains to aid human intelligence, guide us, configure action).
Right now, it’s actionable data that is terraforming industry.
Take Medicine for example.
Early Warning Data
In the years since we decoded the genome, the scientific project in medicine has been to collect millions of data points from the population on blood, heart pressure, cells and DNA in a bid to pivot healthcare towards more preventative treatment.
This has created a flood of information on our bodies, but as scientists draw connections between DNA and health, doctors are becoming overwhelmed.
There is a deep issue of interpretability.
We could feed this data to algorithms. Dr Watson, for example, is showing signs of outperforming doctors in the diagnosis of obscure conditions. Meanwhile a company such as Enlitic can apply deep learning algorithms to recognise suspicious masses on radiological scans with such success that they have been rapidly adopted into thousands of hospitals.
The issue is that these algorithms have only be trained to do a few very specific tasks. If AI is to address public health, it needs better data.
A company such as Sentrian is developing discrete biosensors that can be placed behind the ear, recognising that for many patients with chronic conditions such as congestive heart failure and COPD, the processes that lead to severe illness starts days before the patient becomes ill – with early research showing that in some people, factors such as heart rate variability, sleep duration and body temperature may be indicators of impending crisis.
This is anticipatory data – good and useful information. It is a much smaller set than actionable data but the more we feed the algorithms, the more they will find.
Biochemist Craig Venter has recently undertaken experiments that sequenced the genomes of thousands of people in order to reconstruct their faces simply from the genetic information – predicting your height, body mass, eye colour, hair colour and texture.
“Whether it’s the shape of your face, or a diseased aorta, or a narrowing of your spinal cord, we want to measure it so we can be able to predict those conditions in the future from the genetic code,” he says.
Goldman Sachs estimates that a new $30bn industry will be born as US healthcare embraces sensor fed IoT.
Terraforming the planet
In this way, all of the technologies that will cross the chasm over the next ten years will be used in the service of collecting and processing genuine information from the atomic rush and jumble of collectable data.
AI will be charged with recognising and anticipating disease. We will engage with it, taking increasing responsibility for our own healthcare, allowing it to guide us on what drugs to take, how much weight to lose, how much drink we can continue to take on board before it begins to colour the extremities of the face.
You can download a free copy of my Artificial Intelligence report here.
The next stage will be to create virtual environments. The tech industry has been building the basic infrastructure for virtual environments for a decade – headsets that can fake the visual system, server farms to power rich environments, batteries that can keep us immersed.
Right now, the technology titans are focused on exploiting their enormous user bases, computing power and machine learning algorithms to map and collect useful information from a variety of environments…
So Google, Baidu and Apple can look at the automotive industry — recognise the dysfunction of producing human driven machines with 90% down time — and seek to develop an autonomous car, laden with sensors, that can extract useful data about how to navigate an environment with traffic jams, bridges and irrational human drivers.
And Monsanto can seek to advance ‘precision agriculture’ via it’s Climate Corporation subsidiary, with meter-by-meter profiling of fields and micro-climates that help farmers determine when to plant and when to harvest, the appropriate water, fertiliser and insecticide inputs to use, draw on a constant feeding of data from satellite navigated tractors and combines.
And what’s exhilarating is that this actionable data is available to industries right across society.
So far it has enabled the flat footing emergence of the technology titans but at the same time it has enabled companies such as Uber and Airbnb to emerge from nowhere and for highly disruptive start-ups, such as Enlitic.
And we want this.
We want meaningful data as a substitute for intuition and bias. We will be happy to turn over decision making processes such as those involved in driving a car, diagnosing a medical condition or deciding how much hydration a section of a field needs over to machines.
To what extent will we abdicate our decisions to this intelligence?
In the wake of controversy over his experiments, Benjamin Libet spoke of a veto that is available for 200 milliseconds before we decide to act. He even presented a useful decision timeline in Mind Time: The Temporal Factor in Consciousness…
In the next decade, as machine learning algorithms become increasingly independent and sophisticated, constantly fed with anticipatory data from sensors and processed without delay, we will begin to recognise a tension in our decisions between our conscious impulse — at 200 milliseconds — and the decision we finally take.
I believe this delay will get shorter and shorter as we recognise how useful this information is, trusting the network, inviting it into our domestic routines.
Invest in the enabling technologies
For investors, the optimum way to play it all is to look for those vital companies that enable these data-processing networks.
We are talking about the suppliers of specialist sensors, the drones, the miniaturised component cameras, molecular scale chip makers and network threat detection systems.
My base case: just as it was crucial in the last decade for professional investors to understand what was happening in the financial system, the next decade will depend on our ability to understand this fast evolving system.
In the next few years, the Tech Titans – the likes of Alibaba, Amazon, Baidu, Google, Facebook, Tencent – will play the role of serial monopolies, the brainchildren of controlling fathers who have designs on a radical overhaul of almost every industry.
And with the augmented intelligence and data available through the cloud, new layers of technology will be grafted onto the economy, creating a Cambrian explosion in sensory devices.
Companies will use every technology necessary to develop a richer screen on reality, discovering data that will allow them to outcompete rivals and gain control of new platforms.
In prospect: a global machine with universal intelligence that we can all access.
Our immediate interest however, as investors, should be to focus on companies that are producing technology that searches, navigates, stores, processes and protects genuine information about dysfunctional environments.
In the meantime, feel free to sign up to my free weekly briefings on technology.
You’ll receive copy of my Artificial Intelligence report — heralding the biggest winners, losers and surprising ways it will infiltrate our lives in the next year.
Regards,
Michael and Eoin