[imagesource:wikimediacommons]

To keep up with the ever-growing AI revolution, Google has launched its latest experimental search feature on Chrome, Firefox and the Google app browser.

This hasty rollout needs to come with a hefty pinch of salt from the hundreds of millions of users who may be given some heinous advice or information from the fledgling robot.

The new feature, ‘AI Overviews’ saves you from clicking on links by providing a summary of the search results using generative AI – the same technology that powers rival product ChatGPT – which you will also need some salt for.

Asking “how to keep bananas fresh for longer” may generate a useful summary of tips such as storing them in a cool, dark place and away from other fruits like apples, but ask something a little left-field and the results can be disastrous, or even dangerous, per The Conversation.

Google is apparently scrambling to fix these problems one by one, but it remains a PR disaster for the search giant and a challenging game of whack-a-mole.

Image: Google / The Conversation

AI Overviews might be able to give you some really helpful and accurate answers but it can also tell you that “astronauts have met cats on the moon, played with them, and provided care”, while also recommending “you should eat at least one small rock per day” as “rocks are a vital source of minerals and vitamins”, and suggests putting glue in pizza topping.

The fundamental problem seems to be that generative AI tools don’t know what is true, just what is popular.

There is no info on the web about eating rocks because everyone knows that is obviously a terrible idea but Google’s AI found a well-read satirical article from The Onion about eating rocks and took it at face value. The idiot.

Another problem is that AI doesn’t have human morals, ethics, or values, or even the ability to perceive them. So when they’re trained on a large chunk of the web – even if that comes with a fancy name like “reinforcement learning from human feedback” or RLHF – it still reflects some of the biases, conspiracy theories and debased ideas that have managed to spread across the web.

Image: Google / The Conversation

So while Google plays catch-up with OpenAI and Microsoft, pushing the technology out into users’ hands before it is even ready, you might want to do some deep diving outside of the AI answer box.

Google will have to decide if this is all worth the risk of losing the trust that the public has in its search engine’s ability to find (correct) answers to questions.

[source:theconversation]