Priya Raman

About the author

Priya writes about AI and new technology – what it can do, what it can’t, and what it means for the rest of us.

Reflection’s Beam Is a 501-Billion-Parameter Open AI Model Built to Take On China’s Best

U.S. startup Reflection unveiled Beam, a large open-weight AI model aimed at coding and agent tasks. The company says the weights arrive later this month.

ChatGPT Will Start Showing Picture Ads While It Makes Your Images

OpenAI says it will test a new visual ad format in the US later this month, starting with ads that appear while ChatGPT generates images. Here’s what we know.

An AI Just Crushed the World’s Best Stratego Player — on a Budget of a Few Thousand Dollars

A new AI called Ataraxos beat the strongest human Stratego player 15–1 with four draws. The bigger story: it was trained on 16 GPUs in about a week.

Free Gemini Users Will Be Limited to Google’s Smallest Model Starting Oct. 9

Google is reshuffling which Gemini models each plan can use. From Oct. 9, the free tier gets Flash-Lite only, and usage limits will reset every five hours up to a weekly cap.

A Light-Powered AI Can Screen Deepfake Videos With Almost 98% Accuracy

UCLA engineers built a detector that does most of its thinking with light instead of electricity – and checks 15 videos at once.

AI Chatbots Give a Narrower Range of Information Than a Google Search, Study Finds

Researchers tested 27 AI models and found every one returned less varied information than a plain web search, raising concerns about “knowledge collapse.”

Why AI Chatbots Make Things Up — and How to Catch Them When They Do

AI “hallucinations” aren’t random glitches. Here’s what researchers say causes chatbots to state false things with confidence, and a practical checklist for catching errors before they cause trouble.
12Page 2 of 2

Categories

BusinessHow-ToNewsOpinionTech & AIWellness