Attendees have come to expect a certain level of technology expertise and insight from a Strata Data Conference. And the recent one in San Francisco didn’t disappoint. The program now covers the entire range and ecosystem of big data tools and technologies. All the usual heavy hitters – IBM, Google, AMD, and Amazon – were there, and the sessions ranged from artificial intelligence (AI) to the omniscient cyber conflict and to how Levi’s built their own data science team. That’s just a brief sampling of the many topics discussed, but it gives you a hint of the range of subjects covered and what people in the industry are devoting their time to address. There was a distinct undertone at the event that data is crucial to building successful AI products and services. Akin to it being the lifeblood of AI.
Overall, I believe five key themes emerged over the three days.
- The rise of the data professional
In truth, I could devote an entire blog post to this issue. The exact definition of a data scientist is now unclear with some Tech & Dev professionals showing a preference for the emerging new role of machine learning engineer while others still prefer data engineer or data scientist. Regardless of the title, Data Science is becoming more pivotal in reimagining business processes, integrating new policies and replacing the old ones, and supplying key insights into successfully transforming a business.
- AI is only as good as the data it receives
Artificial Intelligence needs great data to make it work well. From infrastructure to insights and real-time pipelines to appliance ingestion, your data will make a huge difference in your future AI product or service. There is still a sense that although everyone agrees AI is an important area to build products and services around, many are still confused as to what AI can deliver for their business and how they can implement it. We heard that one of the critical blockers is that 81% do not understand the data required for AI, and 80% of the data is either inaccessible, untrusted or unanalyzed. Again, the lack of relevant skills is cited as another reason or at least a challenge to implementation.
- Bots are getting smarter
Lauren Kunze of Pandorabots gave a provocative and inspiring talk about how bots are becoming much more intelligent, and are much more sophisticated than just those disrupting webpage helpers. Again, this is about bringing parts of the data ecosystem together to make these bots more intelligent and to demonstrate that they are learning from their interactions with humans. The question is now what bot logs tell about human interaction with bots?
- Social + Data: Problem or blessing?
Data and social media together are huge and pose either a potential problem or a great solution. The ethical issues are ones we will all grapple with for a while as we figure out the good, bad, right and wrong of these two areas colliding more in the future.
- The rise of cyberconflict
David Sanger of the New York Times provided some compelling evidence of how cyberconflict represents a new era of data warfare and one that may be the most crucial area where data will dramatically affect peoples’ lives. Can AI and cryptography help? We also heard a lot about the US regulatory scene and a new sharper focus on the issue of ethics and cybercrime. Cyberconflict is an area that is going to require more vigilance and collaboration between data scientists and security professionals as they start to build, protect and deliver safe data products, services or tools for gathering information. I’m thinking about the impact of this particularly for future voting scenarios.
Mike Hendrickson is the VP, Technology & Developer Products at Skillsoft.