Google saw fit to bestow a free ticket to a member of the Council to attend the Strata Conference (https://conferences.oreilly.com/strata) in New York City. The Strata Conference is all about Artificial Intelligence, Machine Learning and Data Science, as well as some attendant topics. Size-wise, this conference is on a par with LegalTech and the International Legal Technology Association (“ILTA”) Conference . However that is about where the similarity ends. As far as I could tell, I was the only attorney in attendance. They were curious to know why I was there, but when I described e-discovery, data breach situations and the like — they got it. A very welcoming lot. Here are some observations:
1. As with the legal conventions, AI was all the rage. Not surprising, however, because the Strata conference is all about data, so Data Science (aka “Big Data”) plays an more important role. AI is used to analyze Big Data. There is just so much data that only a machine could cope with it.
2. Open source software applications dominate the field. In fact, there was one presentation entitled “Commercial Software in an Increasingly Open Source Ecosystem.” Large companies now insist on open source applications (for a variety of reasons). So much so that when they go looking for solutions, they turn to the open source versions first. Even Microsoft was touting how well you could run Linux and other open source applications on Azure. Microsoft knows that it has a credibility problem in this area, but they are truly making an effort to make amends with the open source community (and by extension, corporate America).
3. AI and Data Science are quickly being institutionalized in corporate America. Corporate networks are being modified to capture company data for use in AI-based applications (which have an insatiable thirst for data).
4. The pace of AI development is exponential, and that pace won’t slow down anytime soon. Indeed, the pace of infrastructure modifications to take advantage of AI development will ensure that that exponential rate of GPU (graphic processor units) growth continues for the near term.
5. Moore’s law is officially dead — kind of. The current rate of growth of *CPU* capability is 1.1:1, rather than the 1.5:1 during the heyday of Moore’s law. Don’t despair. The growth rate for *GPU’s* (favored by AI applications) is currently 1.5:1. Moore’s law isn’t quite dead yet, but it has shifted a bit.
6. AI is getting easier to develop and use. Software is being developed to abstract the process of creating AI. This abstraction process is intended to insulate normal people from the nitty-gritty of developing AI. I saw one Microsoft engineer build an AI-based chat bot in less than 5 minutes (he timed himself, right in front of us). The point is, AI is getting easier to develop and use all the time. Soon, even a lawyer will be able to do it. What this portends is hard to fathom, but people would be wise to monitor the developments. Even better, AI has now hit the “hobby” level, in that the tools to start developing AI are free (from a software royalty standpoint) and the hardware costs to do it (beyond a snail’s pace) are modest.