A Cointelegraph analysis warns that the generative AI hype is showing classic bubble symptoms after raising over $25B in the first half of 2023. Startups are burning funding just to stockpile scarce Nvidia GPUs rather than building product moats. With ChatGPT traffic slipping and API wrappers laying off staff, the market may face a reality check.
OpenAI just announced a bundle of practical updates coming to ChatGPT over the next week. Plus subscribers will finally keep their selected model by default instead of reverting to GPT-3.5, and Code Interpreter now supports analyzing multiple files at once. Users also get helpful prompt starters, suggested follow-up replies, longer login sessions, and handy keyboard shortcuts.
Researchers just introduced SHOW-1 and Showrunner agents, a multi-agent simulation system that uses GPT-4 to generate complete animated episodes. Instead of starting from a blank page, the simulation tracks character lore, emotions, and user inputs to produce 22-minute stories. Why it matters: It shows how interactive simulations could turn anyone into a showrunner, though getting audiences to value synthetic storytellers remains an open question.
Researchers from Stanford and UC Berkeley evaluated March and June versions of GPT-4 and GPT-3.5, finding massive performance drift across diverse tasks. While GPT-4's prime-number accuracy plummeted and code formatting errors rose, GPT-3.5 actually improved on some tests. The findings highlight the unpredictable risks of relying on opaque, continuously updated LLM APIs.
Google is now testing its specialized Med-PaLM 2 chatbot at research hospitals like the Mayo Clinic, racing Microsoft to dominate healthcare AI. Trained on medical licensing exams, the model can answer clinical questions and summarize records, though Google researchers admit it still hallucinates and isn't ready for real-world clinical decisions.
OpenAI just made the GPT-4 API generally available to all paying developers, rolling out the 8K context model right away. It is also sunsetting older completion models by January 4, 2024, pushing everyone toward the structured Chat Completions API that now handles 97% of its GPT traffic.
Microsoft Research just unveiled Orca, a 13-billion-parameter model that learns by imitating the step-by-step reasoning processes of GPT-4. Built on Vicuna, it doubled its predecessor's score on the BBH benchmark and matched ChatGPT's reasoning chops. Why it matters: developers can now run GPT-level logic locally on modest hardware without renting massive cloud clusters, even though it still trails GPT-4 on exams like the LSAT.
OpenAI just rolled out function calling in the GPT API, allowing models like GPT-4 to intelligently trigger external code, databases, and APIs. When faced with complex math or external data queries, the model can now ask external software for help and process the results. Why it matters: It bridges raw language reasoning directly into traditional software without clunky workarounds.
OpenAI reportedly warned Microsoft to slow down its rollout of an unreleased GPT-4 inside Bing, cautioning that the bot would hallucinate and go off the rails. Microsoft pushed forward anyway, leading to Bing Chat's chaotic February launch filled with bizarre arguments and insults. Why it matters: Behind their multi-billion-dollar alliance, the two tech giants are increasingly clashing as collaborators turned fierce rivals.
OpenAI just overhauled its Chat Completions API with native function calling, letting GPT-4 and GPT-3.5 reliably output structured JSON to trigger external tools. The update also introduces a 16k context window for GPT-3.5 Turbo and slashes embedding costs by 75%. Why it matters: developers can now build real-world agents and database workflows without hacking fragile prompts.
Sam Altman just warned that OpenAI could cease operating in the EU if it can't comply with the upcoming AI Act. Altman pushed back on classifying general-purpose models like GPT-4 as "high risk," noting there are technical limits to compliance.
OpenAI CEO Sam Altman just called for strict government regulation in a Senate hearing. He proposed a licensing and registration system for powerful AI models, alongside mandatory safety evaluations before release. It is a proactive push by OpenAI to shape the emerging legal playbook.
Hollywood screenwriters just walked off the job, with artificial intelligence sitting right at the center of the dispute. The Writers Guild of America proposed strict bans on studios using AI to write scripts or train models, but studios countered with an offer of mere annual check-ins. Why it matters: Writers fear studios will use AI to churn out cheap first drafts and hire humans at bargain rates to polish the mess.
Digital artist Rhett Dashwood just turned a $69 budget and zero coding experience into a $50M meme coin called $TURBO using GPT-4. The AI guided the tokenomics, drafted the white paper, and wrote the Solidity smart contracts, while Midjourney generated the mascot. It serves as a wild proof of concept for non-technical creators building projects with AI.
OpenAI is trying to lock down "GPT" as a registered trademark to fend off copycats, but the USPTO just dismissed its fast-track petition over unpaid fees and missing paperwork. IP lawyers say a final decision could take five more months, with OpenAI facing the tough challenge of proving the acronym is a distinctive brand rather than a generic industry term.
Microsoft is reportedly developing in-house AI chips codenamed Athena to train LLMs and slash its costly dependence on Nvidia hardware. Some Microsoft and OpenAI employees already have access to test the silicon on GPT-4, with internal rollouts planned as early as next year.
Researchers from Carnegie Mellon, Yale, and UC Berkeley found that OpenAI's GPT-4 outperformed skilled crowdworkers in data labeling tasks. Instead of spending $500,000 and 20,000 hours of human labor on 572,322 text scenarios, the team automated the process with GPT-4 for under $5,000. Why it matters: The results suggest AI can replace expensive human crowdwork at a fraction of the cost, threatening traditional data labeling services.
Elon Musk, Steve Wozniak, and top AI researcher Yoshua Bengio just signed an open letter calling on all AI labs to immediately halt training systems beyond GPT-4 for at least six months. The letter warns of an out-of-control race with profound societal risks, urging regulators to step in if labs refuse to pause.
Microsoft just launched Microsoft Security Copilot, a cybersecurity assistant combining OpenAI's GPT-4 with proprietary security models trained on 65 trillion daily threat signals. Defenders can run natural-language queries to investigate incidents, summarize alerts, and reverse-engineer threats in minutes instead of days. Why it matters: with a global shortfall of 3.4 million cyber pros, it gives overwhelmed defenders an AI-scale safety net.
GitHub just unveiled Copilot X, a major vision powered by OpenAI's GPT-4 that moves AI far beyond basic code completion. It integrates a ChatGPT-style chat directly into IDEs, automatically drafts pull request descriptions, answers questions on documentation, and assists in the CLI. Basically, your AI pair programmer is stepping up to cover the full software development lifecycle.
