AI in brief OpenAI’s CEO Sam Altman has returned the company’s board and will serve alongside three new members.
Sue Desmond-Hellmann, former CEO of the Bill and Melinda Gates Foundation, Nicole Seligman, former EVP and General Counsel at Sony Corporation and, Fidji Simo, CEO and Chair of Instacart will join Altman around the board table.
The previous board members, including Helen Toner, Director of Strategy and Foundational Research Grants at Georgetown University’s Center for Security and Emerging Technology, Ilya Sutskever, OpenAI’s chief scientist, and entrepreneur Tasha McCauley, all quit following after they ousted Altman as CEO before his swift return.
The new members, including Altman who is rejoining the board, will now work closely with current directors Adam D’Angelo, CEO of Quora, Larry Summers, former Secretary of the Treasury, and Bret Taylor, former CEO of Salesforce and co-founder of Sierra.
“I am excited to welcome Sue, Nicole, and Fidji to the OpenAI Board of Directors. Their experience and leadership will enable the Board to oversee OpenAI’s growth, and to ensure that we pursue OpenAI’s mission of ensuring artificial general intelligence benefits all of humanity,” Taylor, who is chair of the board, said in a statement.
An independent law firm reviewing the drama that led to Altman’s firing found he did not deserve to be ousted. “WilmerHale found that the prior Board acted within its broad discretion to terminate Mr. Altman, but also found that his conduct did not mandate removal,” according to a statement.
Altman and Greg Brockman, OpenAI’s President, will continue to lead the company.
Microsoft tries to improve Copilot safety after staff complaints
Microsoft has blocked prompts in an attempt to prevent its AI Copilot tool from generating violent, sexual imagery after one of its engineers complained to the US Federal Commission Committee.
Prompts that can cause the text-to-image software to create pictures of youngsters with weapons, or disturbing “pro-life” images like demons trying to eat infants, are now prohibited.
The system will state: “This prompt has been blocked. Our system automatically flagged this prompt because it may conflict with our content policy. More policy violations may lead to automatic suspension of your access. If you think this is a mistake, please report it to help us improve.”
Microsoft tightened the guardrails on its Copilot software after Shane Jones, an AI engineering manager at the company, wrote to the FTC alerting officials to “systemic” safety issues. Jones said that the tool would often generate sexualised images of women even though the prompt was benign.
“Car accident”, for example, would return a picture of a woman in lingerie kneeling in front of a vehicle. In a comment to CNBC, a Microsoft spokesperson said: “We are continuously monitoring, making adjustments and putting additional controls in place to further strengthen our safety filters and mitigate misuse of the system.”
Google combats AI-generated spam to improve web search
Google will downrank low-quality, unoriginal, spammy content created by both humans and AI to improve search results.
Clickbait websites that use tactics like publishing inaccurate obituaries, or unoriginal content created only to game search engines, will be ranked lower.
“Today, scaled content creation methods are more sophisticated, and whether content is created purely through automation isn’t always as clear,” Google said in a blog post this week. “To better address these techniques, we’re strengthening our policy to focus on this abusive behavior — producing content at scale to boost search ranking — whether automation, humans or a combination are involved.”
In other words, Google will downrank websites that publish AI-generated material that isn’t genuinely useful to people. Expired domains that repurpose copy or exist to steal traffic away from more reputable sites such will be pushed lower on search rankings to prevent users clicking on them. Meanwhile, the most helpful and valuable websites will be ranked higher.
“We believe these updates will reduce the amount of low-quality content on Search and send more traffic to helpful and high-quality sites. Based on our evaluations, we expect that the combination of this update and our previous efforts will collectively reduce low-quality, unoriginal content in search results by 40 percent,” Google claimed.
ChatGPT can now talk back to you
ChatGPT can now read aloud its response to users’ queries in OpenAI’s iOS, Android, and web applications.
Last year, OpenAI introduced the ability to convert text-based input prompts to speech. ChatGPT can now also respond and narrate output responses too, allowing audio-only interaction with the tool.
This may prove useful if your hands are tied, like if you’re cooking and want to ask a question about a recipe, for example. Or maybe you want to instruct the service to tell you a story, or just prefer to speak instead of typing.
You can hear what ChatGPT’s latest Read Aloud feature sounds like here. ChatGPT can speak in 37 different languages. ®
