Which AI Companies Are the Safest—and Least Safe?

  • 📰 TIME
  • ⏱ Reading Time:
  • 81 sec. here
  • 3 min. at publisher
  • 📊 Quality Score:
  • News: 36%
  • Publisher: 53%

France Nouvelles Nouvelles

France Dernières Nouvelles,France Actualités

A new report graded companies including Meta, Anthropic, and OpenAI on their AI safety measures. Many were found lacking.

and Google DeepMind are grappling with the potential harms of their technology. It paints a worrying picture: flagship models from all the developers in the report were found to have vulnerabilities, and some companies have taken steps to enhance safety, others lag dangerously behind.

The report was published by the Future of Life Institute, a nonprofit that aims to reduce global catastrophic risks. Thecalling for a pause on large-scale AI model training drew unprecedented support from 30,000 signatories, including some of technology's most prominent voices.

“The findings of the AI Safety Index project suggest that although there is a lot of activity at AI companies that goes under the heading of ‘safety,’ it is not yet very effective,” said Stuart Russell, a professor of computer science at University of California, Berkeley and one of the panelists, in a statement. , Facebook’s parent company, and developer of the popular Llama series of AI models, was rated the lowest, scoring a F-grade overall. X.

In particular, the report found that all of the flagship models evaluated were found to be vulnerable to “jailbreaks,” or techniques that override the system guardrails. Moreover, the review panel deemed the current strategies of all companies inadequate for ensuring that hypothetical future AI systems which rival human intelligence remain safe and under human control.

However, other risks are more fundamental to the way AI models are currently produced, and overcoming them will require technical breakthroughs. “None of the current activity provides any kind of quantitative guarantee of safety; nor does it seem possible to provide such guarantees given the current approach to AI via giant black boxes trained on unimaginably vast quantities of data,” Russell said. “And it’s only going to get harder as these AI systems get bigger.

Nous avons résumé cette actualité afin que vous puissiez la lire rapidement. Si l'actualité vous intéresse, vous pouvez lire le texte intégral ici. Lire la suite:

 /  🏆 93. in FR
 

Merci pour votre commentaire. Votre commentaire sera publié après examen.

France Dernières Nouvelles, France Actualités

Similar News:Vous pouvez également lire des articles d'actualité similaires à celui-ci que nous avons collectés auprès d'autres sources d'information.

How TIME and Statista Determined the Best Companies and Colleges for Future Leaders for 2025How TIME and Statista determined the Best Companies and Colleges for Future Leaders for 2025
La source: TIME - 🏆 93. / 53 Lire la suite »

Water companies blocked from using customer cash for 'undeserved' bonusesOfwat says it has stepped in to use its new powers to halt water firms that cannot show bonuses are sufficiently linked to performance from using customer money to fund the payouts.
La source: SkyNews - 🏆 35. / 67 Lire la suite »

Ofwat says 9 water companies must not pay bonuses out of customer billsUK regulator to block executive bonuses at companies including Thames Water because of performance issues
La source: ftenergy - 🏆 47. / 63 Lire la suite »

Are Utility Companies Neglecting Their Duty to the PublicUtility companies prioritize profits over reliable and affordable energy, driven by government policies that incentivize renewable energy sources at the expense of consumers.
La source: OilandEnergy - 🏆 34. / 68 Lire la suite »