AI Weekly: Highlights from VentureBeat’s AI conference Transform
This week, VentureBeat held its largest AI conference in company history. Front and center were startups tackling compelling challenges as well as executives from the largest AI companies in the ecosystem, like Microsoft, Google, Facebook, and Amazon.
With seven stages and over 1,000 people over two days, it’s tough to follow everything that happened, but here are a few highlights.
Set the table
AWS AI VP Swami Sivasubramanian and Facebook’s AI VP Jérôme Pesenti lead some of the largest AI operations on Earth, and both agree that if companies want to scale their AI operations or become AI-first companies, they have to get their data house in order first.
Whether they’re called a chief data officer or something else, the priority should be a data strategy before considering AI for your business, Sivasubramanian said.
“The number one thing I can say there is you want to get your data strategy right, because if you don’t, when you end up hiring a machine learning scientist and you expect them to come and invent amazing new algorithms, the reality is they spend a large percent of their time dealing with data cleanup and data quality setup and so forth. So getting your data strategy right is probably one of the hardest things,” he said.
Leadership from the top is essential
Whether it’s the incorporation of IoT sensors with legacy industries, starting AI initiatives, or responsible and equitable deployment of AI systems, to succeed businesses need buy-in and support from the top.
“The question is, [are you willing to] to invest 12 months of building [an AI] system? … [If you] don’t have top-down support and understanding for bottom-up initiatives, you’re going to fail miserably because there’s going to be … a longer timeline for ROI,” Hypergiant chief strategy officer John Fremont said.
In recognition of this need, in recent months both Landing.ai’s Andrew Ng and Microsoft pushed education initiatives like the AI Business School made especially for business executives.
Compute and the future of AI
Intel VP and CTO of AI products Amir Khosrowshahi and general manager of IoT Jonathan Ballon said new materials could change the way chips are made and further democratize access to compute. By contrast, Facebook VP Jérôme Pesenti talked about Facebook’s 5x growth in AI training compute use, saying it’s “what keeps me up at night.”
Advances in optimization and software will be essential to the future of AI, he said.
OpenAI found that compute necessary for state-of the-art results has grown 10x annually since 2012.
Applied AI ain’t easy
One of the overarching themes of the conference was applied machine learning, or AI in production today. A number of tools have been introduced to make it easier for professionals using AI to collaborate or make interoperability possible, but a fair number of AI projects still fail, IDC reported earlier this week.
Cloudera’s Hilary Mason routinely shares insights into lessons learned by enterprise customers. To avoid sabotage of your own AI project, Mason says managers need to know the limitations of the systems they use and let their team do their jobs. Landing.ai VP of transformation Dongyan Wang suggests businesses start small.
Don’t take a job at a company without clear AI strategy
Google Cloud Platform AI head Andrew Moore urged developers to skip job offers with companies who have trouble articulating their purpose.
“If you find yourself in an organization where they are saying, ‘Hey, we’re going to introduce AI because our competitors are using AI,’ there is a danger they will be using AI without connecting it to a business model,” he said. “I would just walk away from a project that doesn’t know why it’s using AI,” he said.
Acknowledging industry leaders
At Transform this year, VB held its first-ever AI Innovation awards ceremony to honor top applications by companies moving the industry forward and acknowledge noteworthy works.
Among the winners: Bossa Nova Robotics inventory robots and computer vision for outstanding business innovation, Corti’s deep learning for identification of cardiac arrest events during emergency phone calls for outstanding NLP, and the work of Joy Buolamwini, Timnit Gebru, and Inioluwa Raji to highlight gender, age, and race disparity in the most used facial recognition systems in the world.
VB also handed out Women in AI awards, and AI mentor and rising star awards. Attendance at Transform by people who identify as women went from 5% last year to 30% this year.
Check the AI channel this weekend or early next week for a story on responsible deployment of AI and how to avoid ethics washing.
Both the awards and the conference itself are part of VB’s continued work to not just to cover AI news but to also be a convener and bring the disparate AI ecosystem together.
See this link for a look at complete Transform 2019 coverage.
As always, if you come across a story that merits coverage, send news tips to Khari Johnson and Kyle Wiggers — and be sure to bookmark our AI Channel and subscribe to the AI Weekly newsletter.
Thanks for reading,
Senior AI staff writer