October 29, 2019
This blog post was initially published in VentureBeat.
With the AI revolution solidly underway, tech’s top 5 companies are investing huge amounts of money into AI development and AI engineering talent. Meanwhile, VC investments in AI startups are at all time highs, too. And in between the startups and the top 5 are the Fortune 500 companies — the companies with the most data that stand to benefit most from machine learning — who are just not able to compete for talent.
AI talent is either going for huge money at the top companies or for a moonshot at a startup. That means most of the Fortune 500 are being squeezed out of the AI revolution. And if this squeeze continues, we can expect to see the names on the Fortune 500 list change rapidly over the next half-decade.
In the past, the Fortune 500 has been a hotbed of research and development (R&D) investment. In fact, corporate spending on R&D is at an all-time high. That kind of spending has traditionally allowed global companies to fend off competitive threats from startups that may be nimble but lack access to operating capital on a global scale.
When it comes to AI, though, the game has changed. Five companies — Alphabet (Google), Amazon, Apple, Facebook, and Microsoft — have devoted disproportionate amounts of attention (and spend) to AI. They have access to virtually unlimited financial and computing resources, and they have swept up the best talent in a market where top engineers can command massive compensation packages.
At the same time, venture investment in AI has reached record amounts, with $9.3 billion invested last year but the number of deals shrinking. That has concentrated vast amounts of capital in a handful of incredibly well-funded companies that can either afford to pay top dollar for the engineers necessary to drive success in AI or can draw engineering talent by solving problems that are simply too compelling to watch from the sidelines.
This should not be the case. There is a rich set of open-source AI research software available to these companies and their engineering staff, and the industry-specific data necessary to train industry-specific AI applications is already within their databases, which should give the Fortune 500 big competitive advantages. John Deere, for example, has massive amounts of data on machine performance in agriculture — data that could be used to optimize and bring autonomy to machines or data that could be combined with weather data to recommend or automatically change machine settings for any given climate.
Undoubtedly, every Fortune 500 company has talked about AI and likely invested in a small lab or project. But there are limitations that a Fortune 500 company can’t control. For example, those headquartered far away from Silicon Valley or some other tech-centric region may have a more difficult time attracting top-tier engineering talent. There are others that they can control, however. If the salaries they’re offering to AI engineers are too low, or if they haven’t budgeted for enough engineers to support effective development efforts, it’s time for headcount to be re-examined and re-budgeted.
If the Fortune 500 can’t continue some of its promising early AI wins, both they and global consumers will suffer. Achievements like Visa’s AI-powered fraud detection network were born out of big AI investments from industry giants. Likewise Illumina’s world-class efforts in AI-powered genetic research that could unlock countless new disease treatments. There will be fewer wins like these if corporate leaders don’t concentrate on AI projects that matter to their verticals.
If Fortune 500 companies are to catch up, they must invest in AI development, putting AI first much the way they put software first in the past. Innovation centers aren’t enough. Vanity projects bolted onto existing businesses aren’t enough. For AI to improve America’s core businesses, big investment must happen — and now.