Welcome to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• 🧠 Safe Superintelligence

• 🏆 Ranking open source AI models

• 🚢 U.S. Navy integrates AI into underwater drones

• 🇺🇸 11 Presidential Innovation Fellows to focus on AI

Let’s get right into it!

Sign Up | Advertise | Tools

STARTUPS

Safe Superintelligence

Ilya Sutskever, co-founder of OpenAI, played a key role in the brief ousting and return of CEO Sam Altman, then went silent. In May, he announced his departure from OpenAI, revealing plans for a new venture called Safe Superintelligence Inc. This company aims to create safe, powerful AI systems without the immediate pressure of commercializing products. Sutskever, along with co-founders Daniel Gross and Daniel Levy, plans to focus solely on developing "safe superintelligence" with a dedicated team.

Safe Superintelligence will have offices in Palo Alto and Tel Aviv, emphasizing AI safety through deep engineering and avoiding external pressures. The venture, however, is a high-risk gamble, as investors must believe in the potential breakthroughs of a small, specialized team without the prospect of immediate profits. Despite these challenges, Sutskever is confident about raising funds and advancing AI safety.

OPEN SOURCE

Ranking open source AI models

Researchers have found that major tech companies like Meta and Microsoft are labeling their AI models as 'open source' without disclosing key details about the underlying technology. True openness in AI is crucial for transparency and accountability, especially with the upcoming EU Artificial Intelligence Act, which will regulate AI models less strictly if they are truly open source.

Smaller companies tend to be more transparent, while larger firms benefit from claiming openness without full disclosure, a practice known as open-washing. The study highlights the need for clear definitions and full access to data and methodologies to advance AI responsibly and maintain scientific rigor.

MILITARY

U.S. Navy integrates AI into underwater drones

In collaboration with the Defense Innovation Unit (DIU), the US Navy has successfully employed AI to autonomously detect underwater threats, significantly enhancing maritime operations. Led by Alex Campbell, this initiative has halved the time needed to search for under-sea mines by integrating AI algorithms into underwater drones equipped with sonar sensors.

These advancements have streamlined workflows, reduced mission durations, and allowed for rapid remote updates of AI models. The project, which began two years ago, has benefited from partnerships with commercial tech providers, creating a robust AI infrastructure to detect underwater targets.

GOVERNMENT

11 Presidential Innovation Fellows to focus on AI

The latest cohort of Presidential Innovation Fellows will be the first to focus exclusively on artificial intelligence technologies, aligning with President Joe Biden’s October 2023 executive order on AI. Announced by the General Services Administration, the 11 fellows from private sector companies will spend a year understanding AI’s role in public sector operations.

GSA Administrator Robin Carnahan emphasized that this AI-focused cohort will implement values-driven AI aligned with the executive order to harness its benefits, mitigate risks, and improve government service delivery. Their projects include expanding access to justice, minimizing consumer harm, and enhancing clean energy infrastructure. The fellows will serve as senior advisors in various federal agencies, supporting leadership and advancing AI initiatives.