It’s a data driven world, and it’s getting increasingly difficult to accumulate the kind of data required to stay competitive with the giant technology companies. This has caused some concern about the future of artificial intelligence, an industry expected to become a major source of economic growth. Between 2016 and 2025, AI revenues are expected to increase more than 40 fold, reaching nearly $60 billion. In 2016, tech giants alone spent between $20 billion and $30 billion, 10 percent of which was spent on acquisitions.
If AI is primarily data-driven and incumbents have the best data and means of accumulating it, do startups have a chance? Some signs point to yes.
One area that has people optimistic about the industry is machine learning through simulation. When Alphabet’s DeepMind developed AlphaGo, which eventually went on to beat the best Go players in the world, it relied on data from previous games. Eventually, DeepMind learned how to train the system through simulation.
It turned out that AI just needed the rules of the game Go, and then it could teach itself to play. The result was AlphaGo Zero, a new-and-improved version of AlphaGo that trained for three days and went on to beat the original AlphaGo in 100 games, losing none.
For the robot apocalypse enthusiasts in the crowd, this might be a spooky revelation. This is great news for startups looking to challenge the Frightful Five, along with China’s growing AI tech giants. Data is not necessarily the most important ingredient to becoming competitive in AI.
In the future, the tech giants’ biggest advantage in AI may be the same advantage they have in every other industry: money. The costs associated with AI can already be prohibitive. In a survey of health care professionals last quarter, 28 percent said cost was the biggest obstacle to AI adoption. The second biggest concern was interoperability.
The advantages to adoption are real, though. Senior executives expressed as much in a survey in September when 79 percent responded that implementing AI brought new insights and better data analysis. They also said AI made their organizations more creative (74 percent) and helped them make better management decisions (71 percent).
With billions of dollars being spent in the industry and companies already reaping real benefits from AI, startups are not being scared off by data deficits or high costs. AngelList currently lists nearly 4,000 AI startups with an average valuation of $4.9 million.
While the giants clearly continue to have the edge, opportunities abound in the industry. It seems people want AI for everything these days and Google can’t address every niche itself.
Free Email Updates
Get the latest content first.
Congratulations. Welcome to the family.