The US government moves to further restrict tech exports to China
The US government is gearing up to roll out additional sanctions to thwart American chipmakers from bypassing export limitations on…
Researchers jailbreak GPT-4 using low-resource languages
Using low-resource languages (LRL) like Zulu or Scots Gaelic can elicit unsafe responses from GPT-4 despite its alignment guardrails. Researchers…
DAI#8 – AI gets inside your head and resurrects Johnny Cash
Welcome to another edition of the AI news roundup that will inevitably win a Pulitzer. We’re still patiently waiting for…
OpenAI prepares to offer better pricing to developers
OpenAI is reportedly preparing to cut costs for developers and enterprise users. Set to be unveiled next month, these changes…
Simply fine-tuning LLMs can remove alignment guardrails
Commercial LLMs like GPT-3.5 and OpenAI have guardrails to make sure the models are aligned and don’t generate dangerous responses.…
MindEye retrieves and reconstructs images from brain scans
MedARC develops foundational AI models for medicine and their latest model, MindEye, can tell what you’ve been looking at. MindEye…
ASEAN takes a business-friendly approach to AI regulation
The EU has been pushing for global adoption of its AI Act, but ASEAN countries are reportedly opting for a…
Project Green Light uses AI to reduce vehicle emissions
The frustration of having to stop at a red light pales in comparison to its impact on greenhouse emissions. Google’s…
AI could consume equivalent energy as the Netherlands by 2027
A new study forecasts that the energy consumption of the AI industry might equal that of the Netherlands by 2027. …
AI decodes speech from non-invasive brain recordings
Exactly how our brains process and formulate language is still largely a mystery. Researchers at Meta AI found a new…
