.
@FastCompany talks to HAI’s @drfeifei about the technology she loves, the human-focus it needs, and her work championing values like transparency and safety.
SAFETY
-
Feifei Li on AI transparency, safety, and human-centered technology
By
–
-
Tech Leaders Selling AI Utopia While Ignoring Apocalypse Risks
By
–
This whole thread. with these ppl getting ridiculous amounts of $$, airtime & setting policy selling the utopia they'll bring along with the apocalypse they're also worried about (but also need to work on the utopia that might bring about the apocalypse as well).
— @timnitGebru (@dair-community.social/bsky.social) (@timnitGebru) 21 septembre 2023
I just. Can't🤬 https://t.co/yfhoRXiaLNThis whole thread. with these ppl getting ridiculous amounts of $$, airtime & setting policy selling the utopia they'll bring along with the apocalypse they're also worried about (but also need to work on the utopia that might bring about the apocalypse as well). I just. Can't
-
LLM Privacy Concerns and Production Deployment Discussion
By
–
ICYMI: @predibase CEO / Cofounder @devvret_rishi had a chance to sit down with @LouMM
, @kg4gwa & the team at @TWiT to discuss #LLM privacy concerns, how to overcome, and what it takes to successfully put LLMs into production. Check out the full talk: -
Framework Evaluates Language Model Opinion Reflection
By
–
New policy brief: Whose opinions do language models reflect? Our latest policy publication introduces a quantitative framework that allows policymakers to evaluate the behavior of language models to assess what kinds of opinions they reflect. https://
stanford.io/45XT5gD? -
AI Text Generation Accuracy: Why Systems Fail at Simple Tasks
By
–
We are talking about two different problems. I am talking about prompting a system to write “AI is so cool” on a shirt and it instead writes “AII iss so col”
-
Dark Side of AI: DigitalFUTURES Doctoral Consortium Discussion
By
–
My conversation with DigitalFUTURES Doctoral Consortium on The Dark Side of #AI @_DigitalFUTURES
-
Responsible AI perspectives with CXO Spice host
By
–
Delighted to host you on #CXOSpice @cswolf
. Your POV on #ResponsibleAI and practical advice are relevant and timely! -
Ethics of AI: Preventing Unintended Harm from Robots
By
–
The ethics of AI: how to stop your robot cooking your cat https://
theguardian.com/sustainable-bu
siness/2015/jun/23/the-ethics-of-ai-how-to-stop-your-robot-cooking-your-cat
… This is a piece I wrote from 2015. Hard to believe so much has changed and so much has not AT ALL. -
GenAI Tank Man Image Persists on Reddit Despite Google Removal
By
–
Looks like this GenAI image is gone from the top of Google Images, but still easily found via the Reddit post where it was originally posted six months ago. https://
reddit.com/r/midjourney/c
omments/129pcy8/selfie_tank_man_tiananmen_square_1989/?utm_source=share&utm_medium=mweb3x&utm_name=mweb3xcss&utm_term=1&utm_content=share_button
… -
Watermarking AI-Generated Images: Technical Challenges at Scale
By
–
An important example of how watermarking AI-generated & AI-edited images is already badly needed for all online content. Too bad experts say it could be very hard to do with high accuracy and at massive scale…