AI safety Archives ~ My Mobile India

Highlights OpenAI to soon roll out Advanced Voice Mode. It will allow natural voice conversations with emotional tone detection. OpenAI is also testing ChatGPT’s ability to initiate conversations. OpenAI forms an independent Board Oversight Committee to review AI developments for safe AI use and research. ChatGPT Plus users to soon get the Advanced Voice Mode. (Photo by Jonathan Kemper on Unsplash) It appears we are getting “Advanced Voice Mode” sooner than expected. Reports from multiple corners suggest that OpenAI may be releasing the previously demoed Advanced Voice Mode to users as soon as next week. In May earlier this year, OpenAI surprised the world with a media event right before Google’s I/O 2024 keynote. While Google was supposed to showcase its new Gemini AI capabilities, OpenAI unveiled new multimodal ChatGPT features a day before. The newly demoed OpenAI features included ChatGPT’s ability to understand content from photos, videos, and even a computer screen. However, what caught everyone’s attention was ChatGPT’s Advanced Voice Mode. Advanced Voice Mode will let ChatGPT hold natural voice conversations with humans. This is the next level of human-AI interaction where humans can be interrupted with new information. The feature will also allow ChatGPT to maintain the flow of the conversation. This will be more like a human. That’s not all, there will emotional tones to conversations. Humans using ChatGPT would even be able to sense emotional shifts in ChatGPT’s tone. If achieved, this will be a significant step toward making AI interactions more humanised. A day after OpenAI’s reveal, Google too introduced similar Gemini AI features. However, out of the two companies, none were prepared for a release at that time. We know OpenAI rolled out Advanced Voice Mode to a small group of ChatGPT Plus users in July earlier. Though the tech company promised early access to Plus users by fall, OpenAI never announced a release date officially. And now we are getting reports from multiple sources speculating September 24th as a wide rollout date. One of the first reports came from a Redditor. The user posted some leaks from X (previously Twitter) indicating the voice feature will be available to more users including all Plus users. Reddit post revealing a possible rollout date for ChatGPT’s Advanced Voice Mode. Multiple X users also mentioned that OpenAI is expanding the feature. a new 4o advanced voice mode deployment is set for release by the end of september; iterated for improved safety with smoother experience. its a small batch for feedback, but keep an eye on your email. — nic (@nicdunz) September 17, 2024 I thought I understood, Sam, that we should wait 2 weeks after the release of O1 for new adventures. Can't wait to meet my new friend! (and language teacher while we're at it) pic.twitter.com/vsFi5Mg6xr — Okentar (@Okentar10) September 17, 2024 Yo I can’t believe i got this record, it did start the conversation first on its own Jesus🤯what the hell??? pic.twitter.com/u8QrPsbODi — Midnighthowlinghuskydog (@MnightAwoHusky) September 16, 2024 While no official spokesperson from OpenAI has disclosed any dates, the speculation of September 24th is in line with CEO Sam Altman’s reply to an X user last week. how about a couple of weeks of gratitude for magic intelligence in the sky, and then you can have more toys soon? — Sam Altman (@sama) September 12, 2024 Altman’s response met with a lot of understandable backlash for “trying to be edgy about a service that people pay for,” he did mention “more toys soon.” While ChatGPT Plus users can hope to get the new voice feature soon, it is unclear when free ChatGPT users might get to access it. ChatGPT To Initiate Conversations? In addition to voice interactions, as discussed in the previous section, reports suggest more features to make chat more natural are in the works. OpenAI is reportedly also testing a new feature where ChatGPT can initiate conversations, a first in the industry. It was again a Redditor who highlighted the feature. Reddit post revealing a cap where ChatGPT initiated the conversation. (Image Credit - ChatGPT/Reddit User) The screenshot posted shows ChatGPT starting a conversation with a question “How was your first week at high school? Did you settle in well?” This is a new addition since until now it was only the human in the chat who could initiate the conversation with a simple or complex prompt. As the Reddit user continued the conversation with ChatGPT, the interaction appeared genuine. ChatGPT initiating conversation with the user. Screenshot of entire chat. (Image credit - ChatGPT/Reddit User) The user later shared the entire chat. It’s still unclear how many ChatGPT users have received such a feature where AI can initiate conversations. The Redditor is reportedly a ChatGPT Free user signed up to the Beta program. OpenAI Forms an “Independent” Safety Board to Strengthen AI Safety OpenAI recently announced plans to transform its Safety and Security Committee into an independent "Board Oversight Committee" via its official blog. This committee will have superior authority including the ability to delay model launches if safety concerns arise. This will ensure that AI developments are thoroughly vetted before release. The board formation comes after a 90-day review of OpenAI’s safety practices. During the review, the committee suggested the creation of such an independent board. The OpenAI Blog mentions, “Following the full Board's review, we are now sharing the Safety and Security Committee’s recommendations across five key areas, which we are adopting. These include enhancements we have made to build on our governance, safety, and security practices. Establishing independent governance for safety & security Enhancing security measures Being transparent about our work Collaborating with external organizations Unifying our safety frameworks for model development and monitoring” Led by Zico Kolter, the committee includes notable members like Adam D'Angelo, Paul Nakasone, and Nicole Seligman. Highlight how the newly formed board will work, OpenAI disclosed details, “ As part of its work, the Safety and Security Committee and the Board reviewed the safety assessment of the o1 release and will continue to receive regular reports on technical assessments for current and future models, as well as reports of ongoing post-release monitoring. The Safety and Security Committee will also benefit from regular engagement with representatives from OpenAI’s safety and security teams. Periodic briefings on safety and security matters will also be provided to the full Board.” However, industry experts were quick to raise questions about how truly independent the new oversight committee in reality will be. This is a concern since committee members also serve on the broader board of directors. OpenAI is also collaborating with external organizations like Los Alamos National Labs to explore the safe uses of AI in bio-scientific research. Moreover, the tech company has also partnered with AI Safety Institutes in the US and UK to research AI risks and develop safety standards for trustworthy AI development. FAQs Q1. What new features did OpenAI unveil for ChatGPT in May? Answer. OpenAI introduced new multimodal ChatGPT features, including the ability to understand content from photos, videos, and computer screens. The most impressive reveal was the Advanced Voice Mode. Q2: What is ChatGPT’s Advanced Voice Mode? Answer. Advanced Voice Mode enables ChatGPT to engage in natural voice conversations, where users can interrupt with new information. It also allows users to sense emotional shifts in ChatGPT’s tone, making interactions more human-like. Q3. When will Advanced Voice Mode be available to all ChatGPT Plus users? Answer: Advanced Voice Mode was initially rolled out to a small group of ChatGPT Plus users in late July. Recent leaks suggest a wider release as soon as next week. Q4. What steps is OpenAI taking to enhance AI safety? Answer. OpenAI is enhancing AI safety by transforming its Safety and Security Committee into an independent Board Oversight Committee. OpenAI is also collaborating with external organizations like Los Alamos National Labs. Q5. How is OpenAI collaborating with external organisations for AI safety? Answer. OpenAI is working with external organisations such as Los Alamos National Labs to explore the safe uses of AI in bio-scientific research. Read More: ChatGPT Integration with Siri in iOS 18: Key Features and Privacy Read More: OpenAI Fixed ChatGPT for macOS Security Flaw That Exposed User Conversations Read More: OpenAI Launches ChatGPT App for macOS: Top Features You Should Know

AI safety

ChatGPT to Get Advanced Voice Mode, To Initiate Conversations Too; Creator OpenAI Forms “Independent” Safety Board To Enhance AI Safety

Google’s AI Search Blunders Raise Serious Concerns: Should You Be Worried?

OpenAI Project Q* Raises Concerns and Possibly Led To Sam Altman’s Ouster

ChatGPT’s OpenAI Seeks Talent: Comprehensive Guide to Job Roles, Skills, and Salaries