I agree with all of these points. These points may indicate real boundaries to consider in order to mitigate the dark side of AI. Share what goes into training large language models like the one behind ChatGPT, or allow opt-outs for those who don’t want their content to include some of what LLM presents to users and so on. Rules for Built-in Bias. Antitrust laws prevent a small number of giant corporations from forming artificial intelligence cabals that homogenize (and monetize) nearly all the information we receive. and the protection of personal information used by these know-it-all AI products.
But that list also highlights the difficulty of turning a rousing proposal into an actual binding law. A closer look at the White House blueprint points reveals that they apply not just to AI, but to just about any technology. Each one seems to embody a user’s right that has been forever violated. Big tech companies didn’t wait for generative AI to develop unfair algorithms, opaque systems, abusive data practices, and lack of opt-out. It’s a big deal, and the fact that these issues come up in discussions about new technologies only highlights the failure to protect the public from the ill effects of current technology.
At the Senate hearings Mr. Altman spoke to, one senator after another sang the same refrain. Let’s stop messing around with AI because we’ve screwed up with social media regulation. But legislation to curb past abuses has no statute of limitations. When I last checked, billions of people, including nearly everyone in the United States who can poke a smartphone display, are still on social media, bullied, having their privacy violated, and being terrorized. I know there is. Nothing is stopping Congress from getting tougher on these companies and above all, passing a privacy bill.
The fact that Congress has not done so casts serious doubts on the prospects of the AI bill. No wonder some regulators, notably FTC Chairman Rina Khan, are waiting for new legislation. She argues that current law gives her agency sufficient powers to address issues of bias, anti-competitive behavior and privacy violations caused by new AI products.
Meanwhile, when the White House released an update on the AI Bill of Rights this week, it underscored the difficulty of actually enacting new legislation — and the sheer amount of work that still needs to be done. He explained that the Biden administration is working hard to formulate a national AI strategy. But apparently the “national priorities” in that strategy have not yet been finalized.
Now, the White House is asking technology companies, other AI stakeholders, and the public to submit answers to 29 questions about the benefits and risks of AI. Just as a Senate subcommittee asked Mr. Altman and his fellow panelists to suggest a course forward, the administration is asking businesses and the public for ideas. “We will consider whether each comment contains personal stories, experiences with AI systems, technical law, research, policy, scientific material, or other material,” the White House said in its request for information. I promise. (It took a sigh of relief to see that comments from the large language model were not solicited, but I’m willing to bet that GPT-4 will contribute significantly despite this omission. increase.)