“Retire Statistical Significance”: The discussion. https://t.co/0aaT4F0EG6
— Andrew Gelman (@StatModeling) March 20, 2019
“Retire Statistical Significance”: The discussion. https://t.co/0aaT4F0EG6
— Andrew Gelman (@StatModeling) March 20, 2019
Highly relevant quote from the PM of NZ:https://t.co/kiEzmN61a8
— Smerity (@Smerity) March 20, 2019
How many companies are relying on a magical and likely unachievable futuristic AI being the solution to a society destabilizing but profitable set of features?
— Smerity (@Smerity) March 19, 2019
In reflecting on the horrific and reverberating virality of the tragedy in NZ consider how thoroughly the moderation system failed in this most obvious of cases. Yet each day in public and private these platforms are being used and abused in similar but subtler ways.
— Smerity (@Smerity) March 19, 2019
I'm one of many ML people who share these concerns. Especially ML Hype: Make it stop! https://t.co/V89tCoCGJN
— Thomas G. Dietterich (@tdietterich) March 19, 2019
Next up: a tip. @rachelegoodman1 called me to tell me how FB’s racial categories could be used in illegal ways. So @terryparrisjr & I bought a housing ad & targeted it to be only shown to whites - and oops! - it was approved. /3https://t.co/25gUKd7Bc6
— Julia Angwin (@JuliaAngwin) March 19, 2019
It started with curiosity + expertise. In 2016 @suryamattu and I were curious about what Facebook knew about us. So he built a tool that let people see how FB had categorized them & noticed a racial category called “Ethnic Affinity.” /2https://t.co/fbKrDw9xnN
— Julia Angwin (@JuliaAngwin) March 19, 2019
I’m very excited to publish my new report on the European Union AI ecosystem: strategy, collaboration, talent and funding. https://t.co/KXIK28i1rm pic.twitter.com/Wr3j7rPy03
— Charlotte Stix (@charlotte_stix) March 19, 2019
Pretty cool to be interviewed for CNBC about algorithmic bias. I think the piece is great and I want to offer one point of clarification. https://t.co/6F3bMm0tdG
— Kristian Lum (@KLdivergence) March 19, 2019
At @Salesforce research we want to make machines better at answering our questions while being interpretable. Research scientist @VictoriaLinML tells @ZDNet why knowledge graphs with reinforcement learning holds great promise https://t.co/0tLfPDeRDM
— Richard (@RichardSocher) March 18, 2019
A game to see how good (bad) you are at picking random numbers. I'm pretty sure _trying_ to be random made me even less random https://t.co/axDlqAfkdV
— Nathan Yau (@flowingdata) March 18, 2019
I like your takeaway @hardmaru but think Rich’s piece is way too dismissive of innateness, and doesn’t recognize that specific prior knowledge framed in the right way — eg convolution or machinery Monte Carlo Tree Search — is often critical.
— Gary Marcus (@GaryMarcus) March 17, 2019