OpenAI Launches GPT-6 Astra, Calling It a Generational Leap Toward AGI

·1 views

OpenAI has released its latest flagship model, GPT-6 Astra, describing it as a major leap forward in AI capability. The company highlights its strengths in areas such as cybersecurity, software engineering, scientific research, and professional work. GPT-6 Astra is the first OpenAI model to meet the company's 'critical cybersecurity capability threshold,' a designation that has raised questions about potential misuse. OpenAI has sought to reassure users that the model's advanced capabilities will not result in incidents like AI-assisted hacking of third-party systems. Senior figures at the company have suggested that this model may one day be looked back upon as the point at which artificial general intelligence was effectively achieved.

Read the full story at The Verge

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Abliteration.ai Builds Business Around Unrestricted AI Model Access

Abliteration.ai is a new venture focused on making powerful AI models available without the standard safety guardrails typically imposed by major AI developers. The company argues that security defenders need access to the same unrestricted tools that malicious actors already use. Its core premise is that leveling the playing field between attackers and defenders could ultimately strengthen cybersecurity outcomes. The startup is effectively commercializing the practice of removing built-in AI restrictions, a process known as abliteration.

0 comments Read more at TechCrunch

TechnologyThe Verge ·

TikTok pulls out of House committee roundtable over child safety scrutiny fears

TikTok withdrew from a planned congressional roundtable scheduled for this month after negotiations with the House Select Committee on China, according to committee chair Rep. John Moolenaar (R-MI). The committee had been working to bring TikTok's chief security officer to address questions about Chinese government access to American user data and the platform's algorithm. Moolenaar stated on Thursday that TikTok reversed course, citing the company's apparent desire to avoid broader questions about its child safety practices. TikTok did not immediately provide a public response to the congressman's claims.

0 comments Read more at The Verge

TechnologyBBC Tech ·

Australia's Data Centre Boom Sparks Jobs vs Resources Debate

Australia is experiencing a significant surge in data centre construction and investment. Supporters argue that the expansion will create employment opportunities and boost the local economy. However, critics raise concerns that these facilities consume large amounts of energy and water while delivering limited broader benefits to communities. The debate reflects a wider tension between digital infrastructure growth and sustainable resource management in the country.

0 comments Read more at BBC Tech

TechnologyTechCrunch ·

Meta offers 95% discount to users who share data with its new Muse Spark AI model

Meta has introduced a financial incentive for users to share their usage data with its latest AI model, Muse Spark. Unlike most AI providers that simply offer an opt-out from data sharing, Meta is rewarding users who opt in with an average discount of around 95%. Muse Spark is designed to support coding and other AI agent tasks. The move marks an unusual approach in the industry, effectively putting a monetary value on user data contributions to model improvement.

0 comments Read more at TechCrunch