 |
|
|
|||||
OpenAI is rolling back a recent update to GPT-4o, the default model that powers ChatGPT, following complaints from users that it made the chat bot act like a weirdo. "The last couple of GPT-4o updates have made the personality too sycophant-y and annoying (even though there are some very good parts of it), and we are working on fixes asap, some today and some this week," said OpenAI CEO Sam Altman in a X post spotted by TechCrunch. As of midday Tuesday, Altman said ChatGPT was running on an older, less sycophantic version of GPT-4o for all free users. The company hopes to get paid users back on an older release of the model by later today. "We're working on additional fixes to model personality and will share more in the coming days," Altman said, adding OpenAI would share more information about what went wrong "at some point." OpenAI released the new GPT-4o late last week. By the weekend, people began noticing ChatGPT was being overly agreeable and verbose in its praise. As you can see from the X post below, often that praise was also inappropriate and strange. When is OpenAI pulling the plug on the new GPT-4o ?This is the most misaligned model released to date by anyone.This is OpenAI's Gemini image disaster moment.image credit : r/u/Trevor050 pic.twitter.com/kNcdnEYMDq AshutoshShrivastava (@ai_for_success) April 27, 2025 Improving the emotional intelligence of its models, in so far as an algorithm can posses the trait, has been a recent focus for OpenAI. For instance, with GPT-4.5, the company said the model was better at responding with warmth and understanding than its previous systems. In trying to bring that same capability to the more affordable GPT-4o, it seems OpenAI got something wrong. This article originally appeared on Engadget at https://www.engadget.com/ai/openai-rolls-back-update-that-made-chatgpt-an-ass-kissing-weirdo-203056185.html?src=rss
Category:
Marketing and Advertising
Firefox now lets you organize your tabs. Four years after its biggest rivals launched tab groups, Mozilla published a nearly 1,000-word blog post recounting the feature's long road from user requests to launch. (Consider skipping it if you dont like long-winded acceptance speeches.) "What happens when 4,500 people ask for the same feature?" the company asked rhetorically. "At Firefox, we build it." Of course, those users may have requested tab groups partly because Firefox was the only major browser without them. Chrome, Safari and Edge launched tab groups in 2021. Hell, Vivaldi has had them since 2016. Tardiness aside, Firefox users will welcome the chance to tidy up the clutter. The feature lets you drag and drop tabs into groups and label them by name or color. Mozilla says tab groups are on-device and never uploaded to the cloud. "Tab groups aren't just about decluttering," Firefox product manager Stefan Smagula said. "It's about reclaiming your flow and finding focus again." Up next for Firefox tabs: The tech industry's favorite buzzword. Mozilla is testing smart tab groups, powered by AI, which suggest names and groups based on your open tabs.This article originally appeared on Engadget at https://www.engadget.com/computing/firefox-finally-adds-tab-groups-195130482.html?src=rss
Category:
Marketing and Advertising
As Metas first-ever generative AI conference gets underway, the company is also previewing a significant update on its plans to bring AI features to WhatsApp chats. Buried in its LlamaCon updates, the company shared that its working on something called Private Processing, which will allow users to take advantage of generative AI capabilities within WhatsApp without eroding its privacy features. According to Meta, Private Processing is an optional capability that will enable people to leverage AI capabilities for things like summarizing unread messages or refining them, while keeping messages private. WhatsApp, of course, is known for its strong privacy protections and end-to-end encryption. That would seem incompatible with cloud-based AI features like Meta AI. But Private Processing will essentially allow Meta to do both. Meta has shared more details about how it will accomplish this over on its engineering blog but, as Wired points out, its a similar model as Apples Private Cloud Compute (which allows the iPhone maker to implement Apple AI without sending all your data to the cloud). Heres how Meta describes its approach. Were excited to share an initial overview of Private Processing, a new technology weve built to support peoples needs and aspirations to leverage AI in a secure and privacy-preserving way. This confidential computing infrastructure, built on top of a Trusted Execution Environment (TEE), will make it possible for people to direct AI to process their requests like summarizing unread WhatsApp threads or getting writing suggestions in our secure and private cloud environment. In other words, Private Processing will allow users to leverage powerful AI features, while preserving WhatsApps core privacy promise, ensuring no one except you and the people youre talking to can access or share your personal messages, not even Meta or WhatsApp. The company seems well-aware such a plan will likely be met with skepticism. WhatsApp is regularly targeted by bad actors as it is. To address inevitable concerns from the security community, the company says it will allow security researchers and others to audit Private Processing, and will make the technology part of its bug bounty program that rewards people who find security vulnerabilities in its services. Its not clear when generative AI features may actually be available in WhatsApp chats the company describes its announcement today as merely a first look at the technology but it does note that Private Processing and similar infrastructure could have use cases beyond its messaging app.This article originally appeared on Engadget at https://www.engadget.com/social-media/meta-has-a-plan-to-bring-ai-to-whatsapp-chats-without-breaking-privacy-193556026.html?src=rss
Category:
Marketing and Advertising
All news |
||||||||||||||||||
|
||||||||||||||||||