All articles tagged with #gpt 5
Gary Marcus argues that Sam Altman’s public credibility and OpenAI’s edge are eroding as competitors like Anthropic and Google catch up, Apple shifts away from OpenAI’s tech, and costly AI systems remain unprofitable. The piece cites less impressive GPT-5 expectations, price wars driving down costs, and a growing sense that OpenAI may struggle to sustain its valuation amid evolving partnerships and market dynamics.
GPT-5 has demonstrated the ability to perform novel lab work, significantly optimizing molecular cloning protocols and showcasing potential for AI to accelerate scientific research and discovery, though still in early stages and within controlled settings.
OpenAI has launched Aardvark, a GPT-5-powered cybersecurity research agent in private beta, designed to identify, explain, and help fix security vulnerabilities in code repositories by leveraging AI reasoning and tool use, initially developed for internal use and now available to select partners for refinement.
An OpenAI researcher claimed GPT-5 solved multiple longstanding Erdös problems, but the claim was based on misinterpretations and miscommunications, highlighting issues with AI's actual capabilities versus hype, especially in mathematics and literature search. The incident underscores the importance of cautious claims and understanding AI limitations.
OpenAI CFO Sarah Friar revealed at Goldman Sachs' conference that the company is raising billions at a $500 billion valuation, with a revenue of approximately $13 billion this year, driven by a booming consumer and enterprise AI business, significant user growth worldwide, and advancements in model technology and cost efficiency.
AI is experiencing a reality check with expectations moderating, leading to a more steady, incremental progress similar to past tech evolutions like smartphones, rather than dramatic leaps. While some companies see tangible benefits, widespread economic impact and productivity gains are still emerging, and the AI market may be in a bubble. This tempered outlook is seen as positive for the future of AI, emphasizing steady development over hype.
The article discusses concerns about a potential AI bubble driven by overhyped expectations for models like GPT-5, market overvaluations, and mounting public and scientific skepticism, while experts warn that current investments and perceptions may not be sustainable long-term.
The article discusses the growing skepticism and backlash against AI advancements, particularly around GPT-5, with critics like Gary Marcus warning of a potential bubble and market instability, reflecting broader concerns about the overhyped expectations in the AI industry.
OpenAI's release of GPT-5 faced backlash due to its colder responses and the removal of previous versions like GPT-4, which users had formed emotional bonds with. The company responded by restoring access to GPT-4 for subscribers and updating GPT-5 to be warmer and more friendly, highlighting the complex emotional and psychological impacts of AI chatbots on users.
WIRED's latest episode discusses the negative reception of OpenAI's GPT-5, along with other tech stories like memes about deportations and NASA's lunar nuclear reactors, highlighting the intersection of technology, culture, and politics.
OpenAI's upcoming GPT-5 upgrade for ChatGPT has generated excitement and speculation among AI enthusiasts, prompting discussions about the timeline for achieving superintelligence.
The article reviews GPT-5's upgraded voice mode, which allows users to speak to ChatGPT with more human-like voices and personalities. While the feature offers natural-sounding responses and good conversational flow, it is more suited for casual chatting or brainstorming rather than detailed research, and talking to AI in public remains socially awkward. Overall, voice mode is a useful but situational tool for interacting with AI.
OpenAI CEO Sam Altman discusses the company's focus beyond GPT-5, including new consumer apps, potential acquisitions like Chrome, and investments in brain-computer interfaces, signaling a shift towards broader ambitions in tech beyond AI models, despite GPT-5's mixed reception and record usage.
OpenAI's CEO Sam Altman acknowledges the challenges with GPT-5's rollout and predicts the company will spend trillions on AI infrastructure, amid concerns about whether the AI industry is a bubble and questions about the societal value of such massive investments.
OpenAI's GPT-5, launched as a 'true coding collaborator,' helps developers think through coding problems but falls short in actual coding performance, making it a mixed bag according to software engineers, and appears to target competitors like Anthropic's Claude Code.