主持人:Elvis Yu and Raj Shroff
RTHK's podcast on the latest tech news
DeepSeek, the artificial intelligence firm based in Hangzhou, took the world by storm with its low-cost but powerful chatbot.
Now there could be another Chinese rival to ChatGPT, which many regard as the world's leading AI model.
Called Manus, its developers say the tool can execute complicated tasks with minimal human input.
In a demonstration video, the tool performed in-depth analysis of a bunch of job candidates through their resumes, while also conducted a comprehensive property research and stocks analysis.
So how well does Manus agaist competitors like OpenAI's Operator and Claude’s Computer Use? What risks should be considered before letting an AI agent perform an action on your behalf?
Listen to what AI consultant Raj Shroff has to say in this episode of Tech Tuesday.
You'll also hear about a leaked video of Sony that hints the future of gaming development.
主持人:Elvis Yu and Raj Shroff
Imagining having an AI that can navigate desktop apps, move cursors, click buttons, and type text just like how you would.
Well that’s what Anthropic’s latest generative AI model, Claude 3.5 Sonnet promises to do as it enables users to instruct the computer to execute tasks which it had never been taught before.
While robotic process automation is nothing new, the AI model removes the need to programme a long chain of commands as users can simply type their requests as if they were talking to a chatbot.
How should we harness the power of the tool and what kind of risks should be considered before we let AI take the reins? Listen to our AI consultant Raj Shroff’s thoughts on the issue
You'll also hear about a recently open-sourced Google watermarking tool designed to identify AI-generated texts, while psychologist Pokky Choi outlines signs parents to identify if their child may be developing an unhealthy attachment to AI, after a 14-year-old boy in the US committed suicide to ‘meet’ his virtual girlfriend.