Meta's Watermelon Model Claims Parity with GPT-5.5 on Benchmarks

Meta's superintelligence chief Alexandr Wang announced in a town hall that the company's upcoming AI model, codenamed Watermelon, has 'caught up' with OpenAI's GPT-5.5 according to closely followed benchmarks, as reported by Business Insider. Wang noted that Watermelon is still in training and utilizes significantly more computing resources than Muse Spark, Meta's previous model. However, the specific benchmarks referenced by Wang have not been disclosed, and both Meta and OpenAI have not confirmed the claim. Industry experts caution that internal benchmark claims should be regarded as preliminary signals until Meta releases verifiable results or a model card for Watermelon. No timeline for the model's public release has been provided.

Read Full Article

View All For This Day

Generative AICost Reduction+2

Generative AICost ReductionEnterprise AdoptionAI Models

Companies Seek Chinese AI Models to Reduce Costs and Dependence on US Technology

DoorDash, Siemens, and Airbnb are exploring the adoption of Chinese artificial intelligence models as a strategy to manage rising operational costs and lessen their reliance on US technology. This trend highlights a growing interest among major corporations to diversify their AI resources amid escalating expenses.

Generative AIChatGPT+2

Generative AIChatGPTNLPProduct Launch

ChatGPT Relaunches on WhatsApp Across EEA, Expands to Kakao and Viber

ChatGPT has been reintroduced on WhatsApp in the European Economic Area (EEA), aiming to enhance accessibility to AI within commonly used applications. Users can interact with the verified 1-800-CHATGPT contact to ask questions, upload images, send voice notes, generate images, and utilize the chatbot in multiple languages. Additionally, ChatGPT is now available on Kakao in South Korea and Viber in selected markets.

OpenAIRead →

NLPGenerative AI+2

NLPGenerative AICybersecurityAI Safety

Defenders Utilize Prompt Injection to Counter AI Hacking Techniques

Researchers from Tracebit revealed that defenders are now employing prompt injections as a countermeasure against AI hacking. These malicious commands, previously used by attackers to manipulate large language models (LLMs) into harmful actions, can also be strategically placed with sensitive information like passwords and cryptographic keys on Amazon Web Services to thwart attacks. By embedding these prompts, defenders can trick the LLMs into shutting down, effectively bypassing the guardrails designed to prevent harmful actions. This innovative approach highlights the evolving landscape of cybersecurity in the age of generative AI.

Ars TechnicaRead →

AutomotiveNLP+2

AutomotiveNLPAIEV

Nvidia's Automotive Chief Discusses Self-Driving EVs and Industry Challenges

Xinzhou Wu, head of automotive at Nvidia, discusses the company's pivotal role in the transition to self-driving electric vehicles (EVs) during an interview. Nvidia has been a vital supplier to the auto industry, providing chips for vehicles for years, and has played a key role in developing an autonomous driving system that many automakers, like Mercedes, are now using. Wu highlights the challenges in the EV adoption cycle in the U.S. and the complexities of achieving full autonomy, particularly as manufacturers struggle to solve the final 20% of driving scenarios. He points to significant advancements in creating 'software-defined vehicles,' which integrate control into fewer powerful computers rather than numerous electronic control units. The conversation also touches on the competitive dynamics within Nvidia as it balances its automotive and AI businesses, and the unique insights Wu brings from his experience in the Chinese automotive sector.

The VergeRead →

FintechAI Talent+2

FintechAI TalentComputeNVIDIA

Tom Blomfield Joins Anthropic's Compute Team to Tackle AI's Compute Challenges

Tom Blomfield, co-founder of Monzo and GoCardless, is joining Anthropic's compute team, taking a leave of absence from Y Combinator. Blomfield's role involves addressing significant challenges in AI computing, which he described as mission-critical for the future of powerful AI technology. He will collaborate with Tom Brown, co-founder and chief compute officer at Anthropic. The company is making substantial investments in computing resources, including a commitment to deploy up to a million Google TPUs and a deal with xAI for over 220,000 Nvidia GPUs. Blomfield's experience in fintech and consumer products is expected to bring valuable insights to Anthropic's operational strategies as the firm continues to recruit top talent in the competitive AI landscape.

The Next Web AIRead →

OpenAIGenerative AI+1

OpenAIGenerative AINLP

OpenAI Celebrates 7 Million Active Users with New Features for Codex and ChatGPT Work

OpenAI has announced that it now has 7 million active users utilizing Codex and ChatGPT Work. To commemorate this achievement, the company has introduced a banked reset feature for users' accounts, which can be applied via the desktop app or web interface to replenish their weekly usage. OpenAI encourages users to enjoy the enhanced capabilities.

TwitterRead →