Stability AI’s brev til det amerikanske senat, der advokerer for åbne modeller i AI-tilsyn

Dig Deeper

Fremmedartet

Oversigt: U.S. Senatet holdt en høring om AI-tilsynets fremtid og betydningen af åbne modeller.

Stability AI fremhævet den afgørende rolle, som åbne modeller spiller i at fremme gennemsigtighed, konkurrenceevne og modstandsdygtighed i AI-udvikling.

Betydning for kundesupport

Diskussion: Åbne modeller muliggør offentlig granskning, øger konkurrencen og fremmer hurtig vedtagelse af AI på tværs af økonomien.

Risikobaserede foranstaltninger foreslås for udbydere af beregningskraft, modeludviklere og applikationsudviklere.

Åbne modeller sikrer retfærdig adgang og repræsentation og tillader forskellige perspektiver og samfund at bidrage til AI-udvikling.

Mitigations foreslås for forskellige risikoprofiler, herunder indholdsgodkendelsesstandarder, datakvalitet, træningsmetoder og mere.

Diskussionen omfatter også politikker for cloud-udbydere, sikkerhedsretningslinjer og krav til applikationsudviklere samt fremme af åbenhed i AI.

EU's fokus på åbne modeller i AI-tilsyn

Opsummering: EU bør også prioritere udviklingen af åbne modeller i AI-tilsyn for at fremme gennemsigtighed, konkurrenceevne og modstandsdygtighed i den digitale økonomi.

Samtidig bør EU også tage hensyn til restriktioner, der kan påvirke global forretningsvækst.

Video beskrivelse

The Importance Of Open Models For A Transparent, Competitive, And Resilient Digital Economy.

In BREAKING NEWS from, yesterday, May 16th, 2023, the United States Senate held a hearing to consider the future of AI oversight. Ahead of the hearing, Stability AI shared a detailed paper emphasizing the importance of open models for a transparent, competitive, and resilient digital economy.

The paper from from CEO Emad Mostaque, Stability AI to the United States Senate highlighted the crucial role of open models in promoting transparency, competition, and resilience in AI development.

In the 15 page document the paper discusses the importance of open models for transparency, competition, and resilience in AI, and the challenges posed by oversight. Stability AI, a company that develops AI technology, focuses on developing open models that are transparent, accessible, and human-centric.

Open models enable fair access and representation, allowing for diverse perspectives and communities to contribute to AI development. However, recent AI models pose challenges for oversight due to their complexity, non-deterministic nature, and potential for harm. The paper advocates for a risk-based approach to oversight that acknowledges the variety of actors in the AI ecosystem and encourages open innovation while ensuring transparency, competition, and national resilience in AI.

Source:
https://stability.ai/s/Statement-from-Stability-AI-to-the-Senate-Judiciary-Subcommittee-on-Privacy-Technology-and-the-Law.pdf
https://stability.ai/blog/stability-ai-letter-us-senate-ai-oversight