Britain's largest ever woodland survey of bats would "simply not have been feasible" without AI, scientists say.

Researchers left devices on poles among the trees of 310 woodlands across Britain, recording 98 million bat calls over three years, which would have taken more than 30 years to analyse without AI.

Some of Britain's rarest and most threatened bats, including the red-listed Barbastelle, have been recorded in the south-west of England, including Somerset, Devon, Cornwall and Dorset.

1 comment

sorted by: hot top controversial new old
[–] 1 point 1 week ago

This sounds like a very focused single purpose ai, I’d be surprised if this had anything to do with LLMs.

Nothing wrong with it, but I think BBC’s framing as a positive ai story feels a bit forced when nobodies had any problem with this kind of ai in the first place

  • source
  • 1 comment in c/bats@lemmy.world view that post
    [–] 1 point 1 week ago*

    For the downvoters; this isn’t about generative AI. It’s not the bullshit corpo models that steal everything from the internet and regurgitate it.

    These are models developed by researchers and scientists based on open datasets specifically for this kind of thing. They don’t require nearly the same amount of resources to train, and no books or artist careers are destroyed in the process.

    BirdNet for example is because a bunch of nerds got together and thought “wouldn’t it be cool if we could analyse the bird population in the area based on the sounds.” It’s actually really neat.

    As a bat enthusiast I’m glad that there are models for them too.

  • source