20/20 Vision: AI and Technology in the coming year
Author: Dr. Zunaid Kazi
Chief Technology Officer
AI / NLP / ML
I must admit, sitting down to write my annual prognostication piece gives me a perverse pleasure. I am a father, and like the best of them I do particularly enjoy “Dad Jokes” and puns. I have been waiting all my life to write about my 2020 vision.
Levity aside, I decided to go both deep and wide this year. I will go deep and make five predictions of where AI will take is in 2020 and then I will step back and cast my eye on technology trends in general for the upcoming year.
Speak to me
This year was the year of smart chatbots and increasingly more intelligent conversational agents. This year we also saw tremendous growth in Natural Language understanding. The revolution actually started in October of 2018 when Google released BERT but since then, we have seen a new model bettering the previous one being released like clockwork. Along with almost natural language understanding, we are starting to get a sound understanding of spoken input, accent and grammar independent, and near-normal generation of spoken output. We will increasingly be mobile and will increasingly interact with systems and devices via voice with traditional interfaces such as keyboards and touchpads being used less and less.
Give me a piece of that AI
AI will be increasingly commoditized, and we will see the rise of AaaS (AI as a Service). 2019 saw the beginning of the trend of more and more organizations providing custom AI tools and functionalities that differ from the broader (and often more expensive) offerings by the big players like Amazon, Google, Microsoft, and IBM. These offerings will evolve into true frameworks providing custom analytics in specific domains such as retail, healthcare, etc. Companies will pass on the data to these providers and the providers will return the necessary analytics.
Patient, heal thyself
As healthcare gets more expensive and more complex as people live longer and longer, there is an increased focus on value-based care and outcome-based care. With the increasing availability of user-centric health data (such as from wearables, from genomics, extracted from EMRs), AI will focus more on building models to support precision medicine. Physicians will be increasingly providing treatment and care targeted to the individual and not to the statistical norm. Healthcare will be more proactive, prevention, than reactive thus improving care.
Seeing is believing
Computer vision, whether for good or bad, will be ubiquitous. The drive towards real Industry 4.0 will be accelerated by “all-seeing eyes” predicting assembly line failures, estimating stock, powering JIT logistics, and monitoring etc. without human intervention.
However, the one aspect of computer vision that will impact all of us will be the ever-increasing and almost ubiquitous adoption of facial recognition in all spheres of life. Amongst the positive uses of the technology such as in healthcare (tracking a patient’s medication adherence, predicting some diseases)and retail (no checkout checkout), you also get the negative. Facial recognition will increasingly be used in law enforcement for monitoring, surveilling, and control of citizens. Big Brother will indeed be here.
Driving Miss Daisy
Miss Daisy may not be needing Morgan Freeman (Hoke Colburn) to drive her anywhere. Earlier this year, Elon Musk made a few bold predictions including:
“We will have more than one million robotaxis on the road,” Musk said. “A year from now, we’ll have over a million cars with full self-driving, software… everything.”
And these will be cars with Level 5 autonomy with no geofence implying they can drive unsupervised under any conditions anywhere. We have already had a self-driving truck make an autonomous coast to coast run. The technology is ready – the laws, infrastructure, and public attitudes towards autonomous vehicles will start to follow in 2020.
Block and tackle
In 2018, blockchain went through the “Peak of Inflated Expectations” of the Gartner Hype Cycle. In 2019, it wallowed in the “Trough of Disillusionment”. 2020 will finally be the year blockchain emerges to trundle up the “Slope of Enlightenment”. There will be a wider adoption of the technology, particularly in the finance sector as more and more central banking authorities adopt it in 2020. Real innovation will be driven when more and more players work together to achieve blockchain interoperability through common standards.
5G – truly untethered
Huawei claims that 5G can “support 1,000-fold gains in capacity, connections for at least 100 billion devices and a 10 Gb/s individual user experience of extremely low latency and response times.” AI, telemedicine, smart cities, self-driving cars, Industry 4.0, extended reality and so much more all need extremely high-speed always connected untethered network access. 5G can bring the future closer and the first whiff of the promise of 5G will start to show itself in 2020
Like blockchain, virtual reality (VR) is also going to emerge from the “Trough of Disillusionment” in the guise of Augmented or Mixed Reality. Hot off the heels of ARCore (Android) and ARKit (iOS), AR applications will graduate from games into the real world – augmenting and enhancing your experiences and how you interact with the real world, whether on your glasses or on the windscreen.
How well did I do last year? I will grade myself an A-. I was a bit too optimistic about Deep Reinforcement Learning and too pessimistic about Deep Fakes.
And a real final thought. I may not write again next year. My AI doppelganger may.