Bing hasreleased a blog postpreviewing its new AI - power hunt experience — essentially its interpretation of Google AI Overviews , but better . Instead of Google ’s generated pulley block of schoolbook suggesting that peopleadd non-poisonous glue to their pizzas , Bing ’s plans include bespoke spreads of information , including links to sources and further reading material .
An example from the web log shows the inquiry “ how long can elephant live , ” and the AI - generate results get with the following answer in magnanimous font : “ Up to 70 years in the wild and 88 years in captivity . ” There ’s also a link to the source beside it , along with further excerpts from the same webpage . It then branches into related sections like “ factors that regard life ” and “ oldest known elephant . ” And at the top left , there ’s an forefinger outlining all the matter covered .
All of this is laid out on the left over side of the webpage while the right is dedicated to the traditional search resultant role . Microsoft take note that early data suggest this glide path does n’t affect the identification number of clink to websites and “ hold up a healthy web ecosystem . ”
What ’s exciting about this is the possibility of instantly getting the precise answer to your query without even needing to dawn anything . This experience does exist with Google in some simple situations — if you search “ how old is Biden , ” for illustration , his historic period appears at the top of the results varlet , and it ’s corking . So , imagine if that was the nonpayment for any question you asked — instead of have to click an SEO article and scroll down five paragraph to find your solution , it ’d just be right there at the top of the results Thomas Nelson Page .
As well as being useful for quick answers , the link to further interpretation and related to theme could be neat for inquiry and learning . Microsoft claim that Bing “ interpret the hunt query , ” implying it could also help oneself put up users with the results they want even when their query is n’t very specific . This is something large oral communication model can already do — I ’ve used ChatGPT to help mefind a Jane Austen quote that I could only mistily remember the root of — but it would be capital to have this ability implant right into a search engine .
However , with everyone elsestruggling so much with accuracy , Microsoft must also be facing the same problems . require generated results to include radio link to generator will certainly help people fact - check what they ’re reading , but getting accuracy charge per unit up is essential if the company really wants this feature film to be useful .
Right now , the productive search is still in the early stages , and Microsoft sound out it will just be deliver the feature to a minor percentage of user queries to test and gather feedback . We all feel a flake burned after the trainwreck that is Google AI Overviews , but let ’s hope this next endeavour from Microsoft will do a good caper .
