Thread by @betolive

Live at #io18, human-centered #AI4all thought leader @drfeifei w/ #GoogleBrain's @greg_corrado chatting about the future of #AI for everyone w/ #DianeGreene, pics by @AILaPaz's ambassador @AnitaRui2 & @WTMLaPaz's @wondermony

2/ @drfeifei on #MachineIntellligence history: 3 critical factors converged to brought us the AI boom: advances in hardware, emergence of powerful big data and advances in deep learning and machine learning algorithms

3/ @greg_corrado: renaissance in neural networks & deep learning and recently reinforcement learning leads the re-convergence of disciplines like cognitive science, neuroscience and #AI who have been separated for a while.

4/ @greg_corrado on #AGI: machines can do pattern recognition every bit as well as humans can, but it is a tiny sliver of what goes into something like intelligence. We haven't even scratched the surface of memory, planning, strategy, contingency and even #EQ.

5/ @drfeifei on #AGI: "humanity has a tendency to overestimate short term progress and underestimate long term progress, so eventually we will be achieving things that we can not dream of"

6/ @drfeifei's 1st element to complete human-centered #AI thinking: advancing AI to the next stage, there is a need for AI to be more flexible, nuanced, learn faster, more unsupervised, semi-supervised learning ways, understand emotion, communicate with humans.

7/ @drfeifei's 2nd element is technology and application: this tech is to enhance/augment humans, not to replace them, will replace certain tasks and replace humans out of danger, but the bottom line is we can use #AI to help doctors, disaster relief workers, decision makers.

8/ @drfeifei's 3rd element is combine thinking of #AI as a tech as well as societal impact: bring social scientists, philosophers, law scholars, policy makers, ethicists and historians to study more deeply about #AI social humanistic impact.

9/ @greg_corrado on explainable #AI: unfortunate mythology that deep learning is a black box which it really isn't, there are tools that examine how these systems work and explainability is a big part in terms of making these things available for a bunch of applications.

10/ @drfeifei's 2 major kind of #bias: one is the pipeline of #AI development from bias of data to the outcome of the bias, and the second is the human bias, the people who are developing #AI and the lack of diversity.

11/ @greg_corrado on #AI's future: one of the great opportunities is going to be the development of artificial emotional intelligence that allows products to actually have much more natural and more fluid human interactions.

12/ The recorded "Building the future of artificial intelligence for everyone" session is available at #io18