Google investors worried by latest AI blunders

Stocktake: Issues with AI chatbot Gemini indicate Google is ‘trailing and mis-executing in a fast-moving and high-stakes space’

Shares in Google parent Alphabet slumped in the wake of its latest AI misadventures. Google apologised and temporarily pulled the image-generation feature offered by its AI chatbot Gemini after it produced historical “inaccuracies” – most notably, images of racially-diverse Nazis.

Questionable images weren’t the only issue. Gemini told some users it was “impossible to say” whether Elon Musk’s memes were worse than Adolf Hitler.

Gemini’s take on Hamas and Gaza has also offended many. Currently, Gemini seems to have given up on that front; instead of answering simple questions, it tells users to “try using Google search”.

Analysts were unimpressed. Loop Ventures described it as a “meaningful blunder” indicating Google is “trailing and mis-executing in a fast-moving and high-stakes space”.

READ MORE

Melius Research warned the mere perception that Google is unreliable on AI “isn’t good for business”. Bulls might argue Alphabet shares rallied hard following 2023′s AI-related selloff and that hiccups are to be expected in the AI space.

Alphabet also looks cheap relative to other mega-cap tech stocks. Still, the title of Melius’s note – “Cheap for a reason” – captures investor concern regarding the potential for further AI setbacks.

Proinsias O'Mahony

Proinsias O'Mahony

Proinsias O’Mahony, a contributor to The Irish Times, writes the weekly Stocktake column