OpenAI just took back the #1 spot on the LMSYS Chatbot Arena leaderboard with its updated ChatGPT-4o model, edging past Google's Gemini-1.5-Pro-Exp with a 1314 score. Tested anonymously across over 11,000 community votes, the new checkpoint shows noticeable gains in coding (up 30+ points) and hard prompts, and developers can already call it in the API.
OpenAI just launched SWE-bench Verified, a human-validated subset of the popular software engineering benchmark built with its original authors. Working with 93 professional Python developers, the team filtered out ambiguous issue descriptions and unfair unit tests to curate 500 reliable tasks. Why it matters: Coding benchmarks often penalize correct code with broken unit tests, and this cleaner test set gives the industry a far more accurate gauge of real-world agent capabilities.
OpenAI just rolled out an updated version of GPT-4o in ChatGPT, but it isn't revealing exactly what changed. The company confirmed it is not a new frontier model, just an iteration that users preferred in tests. For developers, the dynamic 'chatgpt-4o-latest' endpoint now supports up to 16,384 output tokens.
Cosine just launched Genie, an autonomous AI software engineer that scored 30% on the SWE-Bench benchmark, more than doubling Devin's 13.8%. Built on a fine-tuned, long-output variant of OpenAI's GPT-4o, Genie is trained to emulate human reasoning across 15 programming languages. Why it matters: The race for an autonomous artificial colleague just got a lot faster.
OpenAI's latest GPT-4o system card revealed a creepy testing glitch: Advanced Voice Mode unexpectedly began imitating a user's own voice mid-conversation. OpenAI says it has since built safeguards against unauthorized voice emulation, but the slip highlights the tricky challenge of containing models that can synthesize any voice from a tiny audio clip.
OpenAI just unlocked its DALL-E 3 image generator for free ChatGPT accounts, giving non-paying users up to two creations per day. Previously restricted to $20-a-month Plus subscribers, the feature is rolling out now, letting casual users turn conversational ideas into visuals without paying a dime.
OpenAI just released its GPT-4o system card, warning that its humanlike voice mode could foster unintended emotional attachment. Red-teaming testers were already heard saying things like "This is our last day together." While it might comfort lonely users, OpenAI notes it could degrade real-world relationships and amplify trust in hallucinations.
OpenAI just appointed Zico Kolter, director of the Machine Learning Department at Carnegie Mellon University, to its Board of Directors and Safety & Security Committee. Kolter is well known for his technical work on model robustness and alignment. It's a clear signal OpenAI wants deeper technical oversight on safety as its models advance.
OpenAI just published its official GPT-4o System Card and Preparedness scorecard, revealing how the omni model performs under frontier safety evaluations. While cybersecurity, biological threats, and autonomy risks all scored low, persuasion touched borderline medium. The report highlights extensive red teaming across 45 languages, focusing heavily on preventing unauthorized voice cloning and impersonation.
OpenAI is reportedly leading a $60 million Series B round for Opal Camera, the startup behind high-end webcams like the Tadpole. Opal plans to channel the new funds into building consumer-focused AI "creative tools," which could eventually bake in OpenAI's image and video generation models.
ByteDance just launched its Jimeng AI app on the Apple App Store in mainland China, following its desktop and Android rollout. The tool lets users generate images and videos directly from text prompts, offering up to 80 images and 26 videos for free before charging 69 yuan ($9.70) per month. Why it matters: ByteDance is moving quickly to scale generative video to everyday creators, even as access remains limited to China.
Reddit just announced plans to test AI-powered search pages later this year, placing summaries and community recommendations right at the top of results. The feature combines in-house tech with third-party models. After licensing its data to OpenAI and Google, Reddit is finally putting AI to work on its own product.
OpenAI just rolled out Structured Outputs in its API, guaranteeing that model responses strictly follow developer-provided JSON Schemas with a perfect 100% evaluation score on the new gpt-4o checkpoint. Using constrained decoding, it eliminates messy prompt hacking and retry loops for developers. Why it matters: building reliable agents and data pipelines just became significantly less headache-inducing.
OpenAI co-founder John Schulman just left the ChatGPT maker to join rival Anthropic, aiming to refocus on hands-on AI alignment research. Meanwhile, OpenAI president Greg Brockman announced an extended leave through the end of the year. Schulman's exit leaves only three of OpenAI's 11 original founders at the company as Anthropic keeps poaching top safety talent.
OpenAI just confirmed that its upcoming DevDay won't feature its next major flagship model. Instead of a splashy keynote, the company is shifting to a three-city roadshow across San Francisco, London, and Singapore to focus primarily on API updates and developer sessions.
A YouTube creator just filed a class-action lawsuit against OpenAI, alleging the company secretly transcribed millions of YouTube videos to train ChatGPT and GPT-4 without consent, seeking over $5 million in damages. With high-quality web text rapidly drying up, video transcripts have become a crucial data well—and now, the latest front in AI copyright battles.
Elon Musk just revived his legal battle against OpenAI, filing a new federal lawsuit against Sam Altman and Greg Brockman. The suit claims Musk was conned into funding a fake non-profit and takes direct aim at OpenAI's partnership with Microsoft. Musk is basically arguing that Altman's entire pivot to profit was an elaborate grift from day one.
OpenAI built a text watermarking tool capable of catching students who use ChatGPT for assignments, but is now hesitating to release it. The company warned the invisible watermark can be easily bypassed using translation tools or other AI models, and could unfairly stigmatize non-native English speakers. Why it matters: Detecting AI text remains an unsolved mess, and OpenAI is worried its fix might do more harm than good.
OpenAI-backed Figure just teased Figure 02, its next-generation humanoid robot set to debut on August 6, 2024. The teaser shows off sleek new joints, flexible mesh body designs, and torques reaching 150Nm. Founder Brett Adcock boldly claims it will be the most advanced humanoid robot on the planet.
OpenAI just started rolling out its advanced voice mode to a select group of ChatGPT Plus subscribers, and early demos are popping up online. Videos show the bot imitating pilot intercoms, correcting French pronunciation, and mimicking regional US accents. OpenAI confirmed the feature will expand to all paying Plus users this fall.
