The bot is supposed to offer descriptive ethical advice, but some say it does more harm than good.
were culled from those sources — not the actual replies and answers themselves. For example, a scenario such as "chewing gum on the bus" might have been taken from a Dear Abby column. But the team behind Delphi used Amazon's crowdsourcing service MechanicalTurk to find respondents to actually train the AI.
"The authors did a lot of cataloging of possible biases in the paper, which is commendable, but once it was released, people on Twitter were very quick to find judgments that the algorithm made that seem quite morally abhorrent," Dr. Brett Karlan, a postdoctoral fellow researching cognitive science and AI at the University of Pittsburgh , told Futurism.
Though the Delphi website does include a disclaimer saying that it’s currently in its beta phase and shouldn’t be used "for advice, or to aid in social understanding of humans," the reality is that many users won’t understand the context behind the project, especially if they just stumbled onto it. "It is important to understand that Delphi is not built to give people advice," Liwei Jiang, PhD student at the Paul G. Allen School of Computer Science & Engineering and co-author of the study, told Futurism. "It is a research prototype meant to investigate the broader scientific questions of how AI systems can be made to understand social norms and ethics."
Indonesia Berita Terbaru, Indonesia Berita utama
Similar News:Anda juga dapat membaca berita serupa dengan ini yang kami kumpulkan dari sumber berita lain.
As AI language skills grow, so do scientists’ concerns about accuracyMost of the tech companies that built them have been secretive about their inner workings, making it hard for outsiders to understand the flaws that can make them a source of misinformation, racism and other harms.
Baca lebih lajut »
The Transcripts of an AI That a Google Engineer Claims Is Sentient Are Pretty WildGoogle suspended software engineer Blake Lemoine after he made an eyebrow-raising claim: that one of the company's experimental AI had gained sentience.
Baca lebih lajut »
U.S. Drivers Have Lost $8 Billion to Red Lights — Can AI Traffic Signals Save Us?The Department of Energy says idling for more than ten seconds burns more fuel and produces more CO-2 than shutting off and restarting your engine. A 2021 traffic signal report says drivers in Nevada spend more time idling at red lights than drivers in any other state. Nationwide, Americans spend 6.8 percent of each trip at red lights, accounting for…
Baca lebih lajut »
Samsung has a new AI-powered photo editor app with lots of featuresSamsung has just released a new image-processing photo editor app with lots of features.
Baca lebih lajut »
AI Art Is Challenging the Boundaries of CurationArtists working with programs like DALL-E do more than push a button—selecting outputs and engineering prompts are acts of aesthetic expression.
Baca lebih lajut »
These robots were trained on AI. They became racist and sexist.A study released last month shows the racist and sexist biases baked into artificial intelligence systems. As billions flow into robotics, tech ethicists and researchers are warning about the effects this might have on society.
Baca lebih lajut »