A robotic from Google has achieved a stage of wide-ranging functionality that hasn’t been seen earlier than. It can converse with you want a chatbot, reply questions on photos and even get the suitable snacks for you from a drawer.
The robotic makes use of a model of a language mannequin referred to as PaLM, which Google researchers first created final 12 months. This is much like the GPT-3 mannequin that powers ChatGPT, however has extra parameters – the variety of variables that may be tweaked to …
Source: www.newscientist.com