Daphne Koller Fireside Chat: International Women & Girls in Science Day

"We need to encourage and support girls and women achieve their full potential as scientific researchers and innovators." — UN Secretary-General, António Guterres

At RE•WORK, we are constantly supporting girls and women working in technology and bring together leading female minds in AI at our ‘Women in’ dinner series, and also as speakers at our summits.

At the recent AI Assistant Summit in San Francisco, 50% of the expert speakers were women, which in a still male dominated industry really highlights the progress women are making to receive equal opportunities in top roles.

This Sunday Feb 11 is International Women and Girls in Science Day, and to celebrate some of the amazing women we have at RE•WORK events, we’re looking at a fantastic discussion that took place this January focusing on research progressions and also the positioning of women in AI.  Co-located with the AI Assistant Summit was the Deep Learning Summit where Daphne Koller, who is a co-founder of Coursera, Professor at Stanford University, a MacArthur Fellowship recipient and currently Chief Computing Officer at Calico Labs, joined us for a fireside chat. Calico Labs is an Alphabet (Google) company that is using advanced technology to understand aging and design interventions that help people lead longer, healthier lives.

Moderating the discussion was Rumman Chowdhury, Senior Principal at Accenture, so we were really honoured to have two such influential women speaking together.

As the discussion got underway and the landscape of AI was explored, Rumman moved towards Daphne’s expertise as we came to the topic of healthcare, before moving onto Daphne’s personal experiences as a women working in the field.

Overview of the fireside chat

As Rumman asked Daphne about the implementation of AI in healthcare, Daphne explained that the number of people +65 years is about to supersede the number of children, which means that the need for medical care is about to increase. ‘This isn’t just happening in the states, it’s a worldwide problem,‘ and a problem that Calico Labs are trying to solve using AI. Further ahead of the game in elderly care are China and Japan due to the larger elderly population of these countries. ‘Think about the single child rule, this hasn’t always been in place, so the population is skewed in one direction’.

Rumman explained that in order to accommodate this need for more care, we’re talking about jobs being automated by AI’, and asked Daphne about the hot topic of media discussion:

‘There are so many articles that must be concerning to medical professionals with headlines like ‘robot beats doctor!’ What are your thoughts for doctors who are worried about this?’

Daphne explained: ‘Before I answer specifically I want to answer about this notion that ‘robots are going take over the world’! We’re nowhere close to a general AI that can do everything that we do and start taking over the world. Computers are getting better and better at clearly defined pattern recognition tasks, we can do problems we couldn’t before, but if you take a programme that can recognise images and ask it to do something with natural language and it’s stumped! We’re nowhere near general AI.

There will be jobs replaced that we didn’t think would go away, for example paralegals as well as medical professionals. In medicine, doctors won’t become obsolete but they'll move to a higher level of the job that’s almost more important. For example in banking, when automated tellers were invented everyone thought bankers would be made redundant but this didn’t happen - people took on different roles like helping with personal banking etc. The same thing will happen with doctors. Because of the ridiculous costs and strains on health services, you currently get 5/10 minutes to explain your issue and try and receive a diagnosis before the doctor moves on. If we can provide basic medical care with AI, we can leave the higher end of medical issues to the experts and people can receive high level medical care from experts at a lower cost.

Daphne went on to say:

I think it’s also a matter of improving the quality of healthcare - when you spend 5 minutes with a radiologist or pathologist they sometimes catch the problem and they sometimes don’t because there’s not enough time to dig into it properly. Machines won’t replace doctors but will help them analyse and provide better care.

It’s difficult because we need a hook to get the general public involved, Rumman explained. But what is that hook? If we say ‘Robot beats human’ people are intrigued, but if you say ‘AI helps human’ people don’t care as much! How can we democratise it better?

Daphne answered, explaining that whilst fake news is far more attractive, people then don’t understand the facts because they’re drawn to the sensationalistic headlines. One of the most important things is for those of us who do understand AI and its potential to help by conveying a realistic narrative. When we get the opportunity to talk to influential people we need to try and hammer in the truths and reduce the hyperbole.

The media is really hungry to deliver everything in a hyperbolic 30 second message, but we need to focus on the ways it’s not just replacing a person or a job, but actually look at the whole picture and the facts. It’s funny, because people imagine a factory of people physically replaced with robots whilst that isn’t what would happen in reality. When the hyperbole radiates you end up in an AI winter and we can’t let that happen again. The value is large but we need to stop scaring people or telling people about things way beyond what technology can offer as they’ll turn around disappointed saying ‘oh well why is there no general AI, you said there would be?’ and people won’t invest in the ideas. Because of ridiculous hype you can end up in a dip similar to what happened with education - we heard that all universities would die and all be online, but we actually got a hybrid with things like Coursera.

Rumman asked Daphne, as one of the only women from the early days of AI. How do we fight against the silicon valley’s shunning of not only women, but also minorities?

Daphne shared that she hasn’t experienced issues as severe as what’s currently in the media, the day to day aspects of being a women being in the male dominated field ‘I absolutely have that problem, I’ll be with a male colleague who will be introduced as professor X and I just get called Daphne.’ Not that it matters, she laughed it off, but it’s still something we should be aware of and that shouldn’t be happening!

‘At Coursera I was introduced to a CO of another company and I said it’s really nice to be connected, my assistant James will be in touch to arrange the meeting, and the response I got was ‘Dear Daphne, please can you give me James’ availability for the meeting’  - they just assumed I was his assistant!

There’s also the problem of being in a situation where you make a suggestion and you’re shunned, and 5 minutes later your male colleague makes the suggestion, and everyone comes out with ‘what a great idea!’ - do you stand up to this, or just let it slide? If you see that happening to your female colleague, stand up to it and say ‘hey, Debbie said that just now!’ - don’t just let it slide!

If you’d like to support women working in AI, join us at our upcoming Women in AI Dinners in London this February 20, Boston this May 22 and San Francisco this June 26.