Van’s Note: Van Bryan here, Jeff Brown’s longtime managing editor. Welcome back to Jeff’s 2021 prediction series. Over the past few days, I’ve been asking Jeff to share some of his biggest predictions for the new year. Remember, you can always catch up on all our published predictions by going here.
Today, Jeff and I are talking about artificial intelligence (AI). Read on to learn the AI milestone that everybody missed in 2020… and discover the breakthrough nobody sees coming in 2021.
Van Bryan: Jeff, let’s talk artificial intelligence today. Once again, this is a topic you’ve been covering for years. Let’s start with your AI prediction for 2020. Can you remind us what it was?
Jeff Brown: When you and I spoke in December 2019, I predicted that an AI would discover a completely new drug therapy without the assistance of a human. In other words, a human would provide the AI with data. But the artificial intelligence would develop the drug therapy all on its own. I predicted that would happen sometime in 2020.
Van: And did it?
Jeff: Absolutely. I nailed that one. And it happened almost right away.
In January 2020, news broke that Japanese pharmaceutical company Sumitomo Dainippon Pharma had announced that it would be taking its new compound, DSP-1181, to human clinical trials for obsessive-compulsive disorder.
But that wasn’t the big news.
The breakthrough was that the compound was entirely invented by an artificial intelligence from a company called Exscientia. The AI was able to generate and sift through tens of millions of possible molecules to determine the best drug candidates.
And it’s worth mentioning that people in the industry didn’t think this would happen. I remember attending a biotechnology conference at MIT in 2019, and people thought this was years away from happening.
And it’s not that they aren’t smart people. They are experts in their field. But most of these professionals are just involved in one specific area. They didn’t see the bigger picture.
But it happened, just as predicted.
Van: And what are the implications of this?
Jeff: You know, occasionally, moments like these come along. They aren’t really covered by the mainstream press. Everybody misses them. But these are the moments I remember.
Years later, we look back on these events and say, “That’s when it all changed.” And that’s what we had with this AI breakthrough.
The “old way” of drug discovery is a manual process. Chemists use some computational models, but it’s usually within a defined set of parameters.
At a high level, we can think of this process of drug discovery as manual “tinkering.” They try one molecular combination. See if it works. Try a different combination. See if it works. It’s slow. It takes a lot of resources. It’s expensive. And it’s inefficient.
But this “new way” of drug discovery changes all that. The AI is unleashed on the drug discovery process. It evaluates billions of different molecules over weeks or days. It creates a shortlist of the molecules with the highest chances of success.
What are the implications?
This is the intersection of artificial intelligence and biotechnology.
It’s going to change everything we think we know about the pharmaceutical and biotechnology industries.
We’ll discover new drug therapies in weeks or even days instead of years. Drugs will be cheaper to develop. And we’ll cure diseases that would have been impossible to treat a few years ago.
And yes, there will be plenty of investment opportunities in this space. In fact, I already made one recommendation to readers of my Exponential Tech Investor service. This is a stock with 10x potential.
[Van’s Note: For paid-up readers of Exponential Tech Investor, read the full report on the stock Jeff just mentioned by going here. Not a subscriber? Go right here to join Jeff’s elite research service.]
Van: I’m guessing you have a prediction along these lines for 2021?
The other breakthrough that happened this year was that Google’s artificial intelligence subsidiary, DeepMind, just announced that its latest AlphaFold software can accurately predict the folding of a protein based solely on its amino acid sequence, with 92.4% accuracy.
That’s important because the way a protein in the human body folds determines if a pharmaceutical will be able to bind to that protein and be effective. In essence, we have to know how these proteins fold if we want a drug to work. Historically, this has been a trial-and-error process.
But now, AlphaFold can predict protein folding with 92.4% accuracy. That’s going to make it much easier to create drugs that do what they’re designed to do.
And here’s the prediction. I predict that AlphaFold will be 98% accurate by the end of 2021. And we will see not just one but several drug therapies produced using this technology.
Again, this was one of those breakthroughs that almost nobody noticed. But it’s going to have profound implications for curing disease. And, of course, there will be plenty of investment opportunities in this space, too.
Van: It’s an exciting prospect. We’ll keep an eye on that in 2021. But I’d also like to ask you specifically about the AI used in self-driving cars. We’ve talked about that multiple times this year in The Bleeding Edge. Any predictions there?
Jeff: Absolutely. Next year will be the year that fully autonomous technology is deployed. It won’t be worldwide. It won’t even be nationwide. But we’ll start to see companies like Tesla demonstrate fully functional self-driving cars that can operate in almost any conditions – in bad weather, on city streets, on gravel roads, and so on.
But my prediction for self-driving vehicles in 2021 has to do with Waymo, Google’s autonomous technology division.
I predict we will see the first major licensing deal between Waymo and an automotive manufacturer.
Van: A lot of people believe that Waymo wants to become a carmaker. You disagree?
Jeff: I do. That’s nonsense. That’s a resource-intensive business with low margins. Google has no interest in being an automotive manufacturer. Its master plan is really to license its self-driving technology so that it can be used by existing carmakers.
So one day soon, we’ll walk into an auto dealership. We’ll see a new car or truck on display in the showroom. And it will have a sign: “Self-driving technology from Waymo.”
Or perhaps it will be a ride-hailing company like Uber or Lyft. They’ll simply license Waymo’s technology and embed it in their cars. Or perhaps it will be a new entrant that specializes in grocery delivery in autonomous vehicles. You simply plop the groceries in the self-driving car, and it heads off to somebody’s front door.
That’s what’s coming down the line. And I predict the first licensing deal will be sometime in the next 12 months.
Van: We’ll certainly track that next year.
One more question. What about AI assistants? You’ve been following that trend as well. Any thoughts on where personal digital assistants could be in the near future?
Jeff: The big development in this space in 2020 came from Amazon. The company developed a semiconductor called Inferentia. This semiconductor was designed specifically for the Echo devices that run the Alexa assistant. This is important because these semiconductors can improve end-to-end latency by 25%.
Why does this matter?
When we ask Alexa a question, there’s a pause before the digital assistant answers. That pause is Alexa “deciphering” our question and sending it to Amazon’s data center. The AI in the data center finds the answer and sends it back to the device. This happens pretty quickly. But it’s not instant.
With this new semiconductor, the process becomes 25% faster. That will allow for more natural conversations.
And this is a growing trend. Technology companies are shifting from generic semiconductors to custom, specialized semiconductors. Again, there are incredible investment opportunities here. We’ll be following that in 2021.
And longer term, these digital assistants will be far more functional than anything we have today. Right now, Amazon’s Alexa, Apple’s Siri, and other digital assistants can only do some basic tasks. They can tell us the weather. They can play music. They can set an alarm. There’s not too much else.
But what if a digital assistant could make dinner reservations for you? What if it could order flowers for your spouse on your anniversary? What if it could do your taxes?
That’s what’s coming with these personal digital assistants. It will be like having a digital secretary. And that day is coming faster than anybody realizes.
Van: Thanks for sharing your insights, Jeff.
Jeff: Happy to do it.
Van’s Note: Be sure you check your inbox tomorrow afternoon for our next edition of The Bleeding Edge. I’ll be sitting down with Jeff to discuss robotics, automation, 3D printing, and the looming “manufacturing renaissance.”
And as Jeff said, artificial intelligence is hitting an inflection point. I suspect 99% of investors will miss this tech megatrend. Will you be one of them? Discover the names of Jeff’s top AI stocks right here.
Like what you’re reading? Send your thoughts to [email protected].