Since the metadata can be removed, its absence doesn’t mean an image is not from ChatGPT or our API. Broader adoption of methods for establishing provenance and encouraging users to look for these signals are steps towards increasing the trustworthiness of digital information.
SAFETY
-
Deepfakes: Platform Dissemination Strategy Critical Factor
By
–
Put another way, a deep fake is neither here nor there unless you have a platform + tools to disseminate it strategically.
-
AI Deepfakes Fuel Fraud: Fake IDs and $25.6M Hong Kong Heist
By
–
AI deepfakes are all over the headlines. Two crazy new ones: -OnlyFake, a site that leverages AI to create near-instant fake IDs. -A $25.6M heist in Hong Kong, with scammers using deepfake tech to convince an employee to transfer funds.
-
4Chan’s Dignif.AI Tool Removes Pornography and Tattoos Automatically
By
–
Did you see http://
Dignif.AI, 4Chan’s new automatic depornification and tattoo removal tool? Who predicted that anti-porn deep fake would take off? -
Integrating Best-in-Class AI Solutions Safely to Production
By
–
“How do you bring together some of the best-in-class ingredients that Google, AWS, or Microsoft are building and safely and securely bring them to production so that you can have a true AI solution?”
— DataRobot (@DataRobot) 5 février 2024
VP of Product Marketing and Field CTO Lisa Aguilar sat down with @TechTarget to… pic.twitter.com/NjIN5RZqKb“How do you bring together some of the best-in-class ingredients that Google, AWS, or Microsoft are building and safely and securely bring them to production so that you can have a true AI solution?” VP of Product Marketing and Field CTO Lisa Aguilar sat down with @TechTarget to
-
No LLM poisoning attacks; Sleeper Agents expected soon
By
–
There haven’t been any serious LLM poisoning attacks shown to exist “in the wild” AFAIK but Sleeper Agents came out Jan 10 so data scraped this year is where you’d start to seem them — give it a few months: https://
arxiv.org/abs/2401.05566 -
Genetic Risk Aversion in Early Development and AI Safety
By
–
it‘s genetic; most babies like to play it safe
-
Reflecting on AI Sentience and the Nature of AI as a Tool
By
–
It’s not controversial whatsoever to say that AI is a tool, like guitar. I think what people are afraid of mostly is AI becoming more sentient and self-conscious in a way that it is not prompted but it generates things out of its own will, for its own ends.
-
Survival as an Emergent Objective in Hostile Environments
By
–
What if the objective is something that isn’t hand-engineered per se, like an agent’s ability to survive in a hostile (perhaps, simulated) environment?
-
Viral Technology Risks: Vaccines, Eradication, and State Security
By
–
The same technologies that are used for creating the virus can be used to create vaccines, but many viruses (eg flu) cannot be eradicated by vaccines, and the risk to every population and economy is too high to allow intentional release of dangerous viruses by state actors.