OpenAI just rolled out an update to GPT-4o in ChatGPT, sharpening its ability to follow complex prompts with multiple requests and handle tricky coding tasks. The model also gets a bump in intuition and creativity—plus OpenAI deliberately tuned it to spit out fewer emojis.
OpenAI just pushed back the release of ChatGPT’s new image generator for free users after demand exploded. CEO Sam Altman admitted the tool was far more popular than anticipated, fueled in part by a viral wave of Studio Ghibli-style creations. For now, access remains exclusive to paid subscribers.
OpenAI just rolled out native image generation inside GPT-4o, making it the new default engine in ChatGPT. It nails in-image text rendering, handles up to 20 distinct objects at once, and maintains character consistency across ongoing chats. Renders can take up to a minute, but the jump in control is huge.
OpenAI just shared three key economic observations on the road to AGI, noting that intelligence costs drop 10x every 12 months while socioeconomic value grows super-exponentially. The post also highlights that AI agents will soon act as junior virtual coworkers, making human agency more vital than ever.
ByteDance just launched Trae, a new AI code editor targeting overseas developers. Built on Microsoft software, it lets programmers generate project-level code using OpenAI's GPT-4o or Anthropic's Claude 3.5 Sonnet—and all AI features are currently free. The app is live on macOS, with a Windows version on the way.
Microsoft just launched Microsoft 365 Copilot Chat, bringing metered AI agents into its free GPT-4o chat experience for commercial users. Employees can now run web searches, analyze files, and trigger custom agents without a full subscription. Why it matters: It gives companies an easy, low-friction on-ramp to deploy workplace agents at scale.
OpenAI just published the full System Card for its o1 and o1-mini reasoning models. By using deliberative alignment during its chain-of-thought, o1 drastically outperforms GPT-4o on jailbreak resistance and reduces hallucinations. The trade-off: smarter models also introduce new risks, meaning OpenAI now has to monitor chain-of-thought deception.
Google just dropped Gemini-exp-1121, an experimental multimodal model that outperforms OpenAI's GPT-4o by roughly 20% across coding, math, and vision benchmarks. It features deeper contextual reasoning and optimized architectures, making it much sharper at turning raw design sketches into functioning code.
OpenAI just rolled out an update to GPT-4o, significantly improving its creative writing chops and file handling performance. Uploaded PDFs and text docs now get deeper, more detailed analysis, while generated text and marketing copy sound noticeably more natural and human-like.
OpenAI just expanded ChatGPT's Advanced Voice Mode to the web, letting users talk to the chatbot directly inside their browser. Powered by GPT-4o's native audio capabilities, the feature supports real-time emotional conversations and nine output voices. It rolls out to paying subscribers first, with free access coming in the next few weeks.
Former Twitter engineers just rolled out Particle, an AI newsreader that breaks down complex stories without stealing publisher traffic. The app offers multi-angle perspectives, an in-story chatbot, and prominent source links to send readers back to the original outlets. It's now free on iOS.
OpenAI just launched Predicted Outputs, a new API capability that slashes latency for gpt-4o and gpt-4o-mini by letting you feed in a reference string. It is tailored for partial rewrites—like updating docs, iterating on prior chat answers, or refactoring code inside an existing file.
OpenAI just open-sourced SimpleQA, a factual accuracy benchmark packed with 4,326 short, single-answer questions built to expose hallucinations. Frontier models like GPT-4o struggle hard, scoring under 40% correct. Why it matters: Measuring how often models fabricate answers—and how persistently they overstate their confidence—is the first concrete step toward fixing AI's chronic guessing problem.
Microsoft just launched general availability for fine-tuning GPT-4o and GPT-4o mini on Azure OpenAI Service. The update also brings Meta's new Llama 3.2 vision models, the Phi-3.5 family, and a new Correction feature to catch hallucinations in real time before users see them. It makes tailor-made enterprise LLMs much easier and safer to deploy.
OpenAI is reportedly on the verge of closing a massive $6.5 billion funding round as soon as next week, but it comes with a strict clause. According to The Wall Street Journal, investors can claw back their money if OpenAI fails to fully convert to a for-profit company within two years. Why it matters: Even with $4 billion in revenue, mounting losses and fierce rivalry are forcing the AI leader to ditch its nonprofit roots for survival.
Biotech firm Genmab just expanded ChatGPT Enterprise across its entire organization, issuing over 2,000 licenses under its "AI Everywhere" initiative. Teams built over 100 custom GPTs to draft clinical trial documents and translate research, reporting an average time savings of 3.5 hours weekly. Why it matters: It shows frontier LLMs moving past generic productivity into complex, regulated biotech workflows.
OpenAI just turned its Safety and Security Committee into an independent board oversight group chaired by CMU's Zico Kolter. Crucially, the independent panel now holds the formal authority to delay major model releases until safety concerns are addressed. It marks a significant structural shift to install an external brake on model deployments.
OpenAI just revealed that ChatGPT reached 200 million weekly active users, doubling its user base from last fall. The surge was driven by new models like GPT-4o and bringing Plus features to free users, while 92% of Fortune 500 companies now use OpenAI's tools.
Plaud just launched the NotePin, a wearable AI device built for a single task: taking notes. Unlike ambitious gadgets like the Humane Ai Pin, it can be worn as a necklace, wristband, or clip to capture meetings and dictate memos, transcribing them with OpenAI's GPT-4o. Why it matters: scaling back from replacing smartphones to nailing note-taking might actually make wearable AI practical.
OpenAI just launched fine-tuning for GPT-4o across all paid developer tiers, giving teams a way to tailor its flagship model with as few as a dozen custom examples. Training costs $25 per million tokens, but organizations get 1M free training tokens daily through September 23.
