How has the AI industry developed so far and what have been the major milestones in the AI industry?
The first thing you have to understand about AI is that it’s not the picture most people have in their heads—robots acting like humans or taking over the world. First, robots are a narrow vision of the “body” in which AI might be present. The reality is that right now, your Amazon Echo, your smartphone, and even your car’s antilock brake system is the “body” for AI. That leads to the second point, once AI is developed, it quickly becomes rather mundane—part of the background of our existence. AI powers every search we make, about half of all stock trades, and the price of airplane tickets and hotel rooms.
[easy-tweet tweet=” AI will become a critical factor in the success and market share of most companies” hashtags=”AI, MarketShare”]
In terms of market value, what is the ultimate potential of the AI market?
Just how lucrative will this industry be to those who lead it? Beyond “huge”; the potential of the AI market is hard to gauge because it will be interwoven into the fabric of so many industries, B2B and B2C products and services. An estimate of 15 billion in direct revenue (software, services, hardware) in 5 years would not surprise me.
In the longer term, AI will become a critical factor in the success and market share of most companies. The difference between the AI haves and have-nots will be akin to the difference between businesses which are online vs. only offline.
Who in your opinion is leading the AI industry and why?
The big leaders are Google, Amazon, Facebook, and Apple. They share three critical characteristics: Huge data sets, proven success in the practical application of Narrow AI, and the commitment of R&D and M&A resources to AI investment.
How do you think the AI industry will play out in the next five years?
The next five years are particularly exciting. AI is like a snowball rolling down a hill and it’s finally big enough that we can see it easily and watch it grow before our eyes. From our perspective, the breakthroughs in education alone will be transformative. Our early tests are pointing to the feasibility of vocabulary acquisition at a speed which is an order of magnitude faster than what is currently available. And Lingvist isn’t the only company already using AI to effectively tailor education around the skills and capabilities of an individual learner. It’s ironic that one of the early successes of artificial intelligence will be to advance human intelligence.
What are the implications of the AI industry becoming something that is controlled by two or three dominant players? How will it influence innovation and product development if AI algorithms are more widely distributed around the world, rather than confined to those aforementioned tech behemoths?
In consumer and commercial applications there is a virtuous cycle in AI; AI feeds on large data sets. If AI can provide a product advantage, which in turn attracts larger user numbers, then it gets better and delivers greater product advantage, and so on. For Lingvist specifically, we have a number of insights now that would not have been possible without the size and activity of our learner base. Among them: it takes about 17 hours to learn 2000 words; a learner’s inherent knowledge before using Lingvist can be predicted fairly easily by testing them on a handful of words — and perhaps most revealing, the actual “forgetting curve” as pioneered by Hermann Ebbinghaus (and used by a number of AI education applications) does not fit Ebbinghaus’ original mathematical model — our data implies a substantially different equation is a much better match.
Because of the feedback loop and competitive advantage granted by AI, there is some danger of consolidation among a handful of players. The barrier to entry for someone who doesn’t possess access to a large data set could become formidable.
However, given the players involved, it’s quite likely that one or more will provide AI as a Service (AIaaS)- Amazon very likely, and quite possibly Google as well. This could actually be a powerful enabler of innovation and diverse applications—similar to the way AWS enabled a host of startup companies by providing cloud based, affordable hosting.
Scott Dodson is a serial entrepreneur and Techstars mentor, and current CPO for Lingvist. Lingvist is an AI-based language learning company headed by CEO and former-CERN nuclear physicist Mait Muntel, PhD and is chaired by Skype co-founder and Google-acquired AI-firm DeepMind investor Jaan Tallinn.