Researchers from the Indian Institute of Technology Bombay and Adobe Research have developed a new method called PTP that can reconstruct an AI's secret system instructions with nearly perfect accuracy by analyzing its generated text. Separately, Cathie Wood of ARK Invest has challenged the prevailing view on open-source AI risks, suggesting that widespread access to open AI will compel companies to invest heavily in proprietary solutions from companies like OpenAI and Anthropic for security. AI
IMPACT A new method could expose AI system instructions, while differing views on open-source AI highlight ongoing debates about its security implications.
RANK_REASON The cluster contains two distinct items: one on a new AI research method and another on an opinion piece about AI risks, neither of which constitutes a primary release or significant industry event.
Read on Mastodon — mastodon.social →
- Adobe Research
- Anthropic
- Ark Invest
- Cathie Wood
- Indian Institute of Technology Bombay
- OpenAI
- PTP method
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
