If you happen to assume are good at understanding how the dangerous guys assume, OpenAI CEO Sam Altman has a brand new “preparedness problem”. The highest ten winners will get $25,000 and may even land a job with OpenAI.
What Occurred: Altman has introduced a brand new “preparedness problem” for folks to estimate simply how properly they understand how dangerous guys assume and may misuse large-language fashions like OpenAI’s ChatGPT.
See Additionally: Microsoft’s Revolutionary Glass Storage Know-how Can Retailer 1.75M Songs On A Coaster-Sized Plate
Altman’s problem is not only for enjoyable, although – the bigger motive behind that is making folks conscious of how malicious actors can misuse AI and large-language fashions in other ways.
Altman’s OpenAI will supply $25,000 in API credit to the highest ten entries to make it much more attention-grabbing. One of the best members may also land a job with OpenAI.
“In case you are good at excited about how the dangerous guys assume, attempt our new preparedness problem targeted on novel insights round mannequin misuse,” Altman mentioned in a publish.
Other than the problem, the Microsoft Corp.-backed MSFT OpenAI has launched a brand new “Preparedness crew” to sort out the security dangers associated to AI professionally.
Why It Issues: Giant-language fashions like OpenAI’s ChatGPT, Dall-E, and others are susceptible to be misused by malicious actors. From spreading misinformation propaganda to extra severe crimes, these frontier AI fashions can do much more harm due to the dimensions they provide.
Security dangers for AI grow to be an much more necessary side to concentrate on on condition that firms like OpenAI, Alphabet Inc.’s GOOG GOOGL Google, Meta Platforms Inc. META are all working in the direction of acquiring AGI, or synthetic common intelligence.
Tech personalities like Elon Musk, Apple co-founder Steve Wozniak, and others have referred to as for a “pause” on AI growth, fearing the misuse of AI by malicious actors.
Whereas Altman, too, has sought regulation of AI, he has not referred to as for a pause but. Nonetheless, the preparedness problem suggests he’s getting ready to regulate the risk earlier than a malicious actor goes rogue and causes widespread havoc.
Picture by way of Shutterstock
Take a look at extra of Benzinga’s Client Tech protection by following this hyperlink.
Learn Subsequent: Mark Zuckerberg Says Your Feed Will One Day Have AI-Generated Content material, Not Pals And Household
