Voices in AI – Episode 110: A Conversation with Didem Un Ates

Voices in AI – Episode 110: A Conversation with Didem Un Ates


About this Episode

On Episode 110 of Voices in AI, Byron speaks with Didem Un Ates, the Senior Director of AI buyer and companion engagement for Microsoft about Artificial Intelligence.

Listen to this episode or learn the complete transcript at

Transcript Excerpt

Byron Reese: This is Voices in AI dropped at you by GigaOm. I’m Byron Reese. Today, my visitor is Didem Un Ates. She is with Microsoft and her title is the Senior Director of AI Customer and Partner Engagement. She’s been there for a number of years. She holds two levels, together with in electrical engineering, from the University of Pennsylvania and he or she has an MBA from Columbia as nicely. She’s becoming a member of us from London. Welcome, to the present, Didem!

Didem Un Ates: Hi, Byron. Thanks for having me.

I all the time like to begin with the identical group of questions, which begins with: What is synthetic intelligence and why precisely is it synthetic? What’s synthetic about it and what’s intelligence for that matter?

Thank you. The method I attempt to clarify it to my prospects, companions and different people like college students at colleges – universities or excessive colleges, is principally: synthetic intelligence is a method of mimicking our mind. Intelligence is smart of issues round us. It’s how we course of the environment, how we make sense of it, make these connections between the previous and the long run and the current, that’s known as basic intelligence. Then we even have particular intelligences, which is all very particular capabilities like object recognition or speech recognition. ‘Artificial’ is making an attempt to imitate this with know-how, with algorithms.

Read More:  Leaked S-1 screenshots show Palantir losing $579M in 2019

Well, it’s fascinating you’re saying the phrase ‘mimic.’ Is that to indicate it’s not precise intelligence? It’s simply doing one thing that may emulate intelligence or do you really assume it’s good?

No, it’s positively good and it’s – in some circumstances, the precise intelligence that I referred to, some name it weak AI, is definitely already smarter than people in these areas. Microsoft really was the primary to surpass human intelligence in speech recognition, translation, object recognition and so on. Yes, a few of these practical areas are already very good and even smarter than people, however the basic AI, or the robust AI as some prefer to name it, is round a 5 yr outdated’s intelligence degree. That’s why I name it ‘mimic.’

When you say – you assume we’re at a 5 yr outdated [level] for basic intelligence, is that actually the case? It appears to me that we now have this one trick that’s been working fairly nicely for some time, which is machine-learning, the place we take a bunch of information in regards to the previous and we research it and we make projections into the long run. That appears to be a extremely – not a really generalized device. There are a number of issues the place the long run’s not just like the previous. The phrase ‘banana’ is claimed the identical method tomorrow and yesterday so it’s a extremely good factor that you may do this.

Read More:  4 months into lockdown, Eventbrite CEO Julia Hartz sees ‘exciting signs of recovery’

Things like creativity and different types of issues we affiliate with basic intelligence, are they even solvable that method? When you say we’re at a 5 yr outdated, meaning perhaps subsequent yr we’ll be at a six yr outdated, at a seven, after which in 15 or 20 years, we’ll be at a teen. Is there a restrict to our one little trick we all know right here and what it’s going to have the ability to do?

These are nice questions, and I feel much like you, Byron, I’m obsessively studying about AI and making an attempt to get completely different views on the consultants, largely on the universities but additionally the trade. To me, once I say – or after we learn in regards to the basic intelligence is across the age of a 5 yr outdated human being proper now, all it means is as we enhance the algorithms round AI and ML, we mimic the human studying – human mind and it’s on the degree of a 5 yr outdated human being. Some after all predict really that basic AI are literally racing to achieve an grownup human thoughts. It’s my private view, not Microsoft’s view by far, however my private view is: sure, AI/ML will attain grownup intelligence, however when it will occur is an enormous query.

Read More:  Self-driving vehicle startup Argo AI completes $2.6B deal with Volkswagen, expands to Europe

To your level about teenager years, some predict that will probably be occurring in 5 years; others are saying it gained’t occur in a century. The common not less than in my studying and analysis is someplace round 15 to 25 years. This is totally my very own – let’s say doing my very own homework. This is kind of critical as a result of it has many implications when it comes to let’s say, automation or impression on society, jobs, adjustments, thrilling issues coming, and likewise a number of integrations that we should always proactively handle when it comes to accountable moral AI, which we’re, as Microsoft, very, very, critical about.

Listen to this episode or learn the complete transcript at


Byron explores points round synthetic intelligence and aware computer systems in his new guide The Fourth Age: Smart Robots, Conscious Computers, and the Future of Humanity.


Add comment