Skip to main content

AI Weekly

AI Concept
By Jed Backhouse
29 February 2024
Digital and Insight
artificial intelligence

This week in AI news has seen a large amount of fallout from recent Google Gemini updates resulting in a new image generation tool being taken offline soon after launch. CEO Sundar Pichai has confronted the topic publicly in a memo this week, while using the opportunity to promise “improved launch processes” in the future.

It’s interesting that his statement in multiple places positions Google’s AI tools as worthy of — or wanting to be worthy of — people’s trust. A clear brand statement being used to heal what was a pretty damaging issue for the tech giant’s AI capabilities, with more stories appearing in recent days across the likes of the BBC and Bloomberg.

Elsewhere, the “AI election” as it’s been dubbed is seeing more stories emerge, as The Guardian asks: is the US prepared for AI’s influence on the election? The article cites recent political deepfake controversies in Slovakia, Indonesia and India, with the view that the US regulatory landscape is currently not ready.

On the other side of the Atlantic, Meta (Facebook, Instagram, Threads and WhatsApp) has spoken publicly about its plans to form a team to tackle deceptive AI content in this year’s EU elections in June – with a large part of their plans being around AI-generated images.

Swedish fintech business Klarna has made headlines after publishing a statement saying its AI assistant is doing the work of 700 full-time agents. The AI assistant, available on the Klarna app, is powered by OpenAI and features a chatbot that has handled 2.3 million customer service chats in 35 languages in its first month alone.

A UK police initiative has taken to the roads, which uses AI to detect whether motorists are using mobile phones or not wearing seatbelts while driving. The technology was first tested and is still in use in Australia, where the number of mobile phone detections have dropped by a factor of six.

Piers Morgan, Nigella Lawson and Oprah Winfrey are the latest celebrities to be caught up in a deepfake scandal – to do with a US influencer’s controversial self-help course. This BBC article includes the videos, originally posted on YouTube, which have now been removed.

And finally, if you (like me) thought the deepfakes weren’t all that convincing in the above article, be prepared for what Alibaba’s new ‘EMO’ tool (short for Emote Portrait Alive) can create. Check out this thread of posts to see Leonardo Di Caprio rapping Eminem and Mona Lisa reciting Shakespeare.