An AI biologist

An AI  biologist
Not a valid attachment ID.
About ten miles off the  coast of California on a foggy October morning, a crane lifts a boxy yellow robot  from the deck of the research vessel Rachel Carson and lowers it into  the murky, metallic-gray waters of Monterey  Bay. The remotely operated  vehicle, packed with cameras and  lights, remains tethered to the ship by  a detachable cable, but  the artificial intelligence has  taken on a  will of its  own.
The robot, called  a MiniROV, broadcasts images of a rarely seen jellyfish  onto a wall of screens lining a cramped control room aboard the ship.  Kakani Katija, a senior engineer at the Monterey Bay Aquarium Research  Institute, lightly  gripped a pair of  levers, maneuvering  the MiniROV closer to the translucent  creature swimming  in a marine  blizzard of organic particles falling to the  sea floor.
MBARI researchers  working with the MiniROV aboard the RV Rachel Carson in late October.  Photo: Rachel Bujalski/Bloomberg
Then  Paul Roberts, a senior electrical  engineer, presses a  button on a laptop and announces,  "We're starting to track the agent."
The "Agent" is an AI program  built into  the MiniROV's control algorithms. It's the first time  it's been deployed in the ocean, allowing the robot to autonomously locate and track marine organisms. Katija  lets go of the  controls and smiles as the robot begins  to chase the square-shaped jellyfish  on its  own, using thrusters to  keep up with the  animal's pace as it  moves away.
To build  a network of  AI-powered ocean-surveillance robots, MBARI researchers are  using citizen scientists to play a game that  can train the machines much faster than a small group of  researchers. The need for speed is clear:  the ocean is the  world's defense against climate  change, and marine organisms play  a vital role in  recycling carbon  from the  atmosphere. Thanks to advances in robotics, underwater  cameras, and sensor technology, researchers have  amassed millions of images of  sea life,  but most  of the creatures remain unknown.  Indeed, classifying them  could take years, limited by the availability of  overworked taxonomies. The jellyfish that  the MiniROV is tracking was first discovered in  1990, but  it wasn't identified until 2003 as a new species, Stellamedusa  ventana. The game for phones and  tablets, called FathomVerse, fills a virtual ocean with images of  sea creatures in their  underwater habitats, stored in a  vast database  called FathomNet. Some  of the photos  feature marine animals whose  identities have been verified by scientists. Others are  AI-tagged organisms or have yet to be  classified.
After players  train as amateur marine biologists, they  embark on missions  by moving along ocean currents  in search of pulsating dots that indicate where marine life has been recorded. Players tap the screen to see the animal and identify  whether it is familiar or  marked as unknown. The game then reveals whether their choices match the consensus of other players or if the creature remains  undecided.
They earn points for correct  classifications, as well as the number of organisms they  distinguish. Players also  receive bonus points  if they correctly  label a previously unidentified life form when  a consensus is  reached.
In the  background, researchers  check the players' consensus  results and compare them  with the  AI classification. Chris Jackett, a research scientist at  the Commonwealth Scientific and Industrial Research  Organization of Australia, says games like FathomVerse  could play  an important role in training AI.
"The human ability to recognize patterns and identify unusual features remains unmatched, and having multiple observers  looking at the same  images helps  create powerful training datasets," says Jackett, who works on AI detection and identification of corals and other marine  species.

[attachment deleted by admin]