Almost all of the latest AI implementations are based on Machine Learning. That means that access to training data is critical.
Andrew Ng – Baidu’s ex Head of AI, and one of the the most influential minds in AI – makes this interesting point on HBR:
“After understanding what AI can and can’t do, the next step for executives is incorporating it into their strategies. That means understanding where value is created and what’s hard to copy. The AI community is remarkably open, with most top researchers publishing and sharing ideas and even open-source code. In this world of open source, the scarce resources are therefore:
- Data. Among leading AI teams, many can likely replicate others’ software in, at most, 1–2 years. But it is exceedingly difficult to get access to someone else’s data. Thus data, rather than software, is the defensible barrier for many businesses.
- Talent. Simply downloading and “applying” open-source software to your data won’t work. AI needs to be customised to your business context and data. This is why there is currently a war for the scarce AI talent that can do this work.”
We’re already seeing that access to data is becoming harder and harder (also driven by regulation like GDPR). And yet, the only way to improve the AI technology is to get access to more, and better data. That is the challenge that every AI technology vendor and solution provider is facing.
There are a few ways to navigate this puzzle:
- Pay for existing data
- Pay humans to generate data
- Offer services (often free) in exchange for data
- Figure out a way to improve the technology, while not owning the data
While counter intuitive, the 4th option is the only real key to the Data Access Puzzle.
This puzzle can only be solved by accessing the data while providing the ultimate control and ownership to the rightful owner – the Enterprise (or the Consumer). This will automatically disqualify vendors who have a poor reputation on Data Privacy, or build their business model on vacuuming data.
Disclosure:
I work with a team of researchers and engineers who strongly believe that this is the only way to offer real AI solutions for the Enterprise. Our secure AI solutions are a testament that this approach works.
