Computers can perform maddeningly long calculations in a matter of milliseconds or open up a whole new online world for us at the click of a button. They’re constantly improving to the point where we’ll soon struggle to discern computer graphics from the real world, but one thing they can’t quite match us humans in is our ability to make snap decisions based on a series of inputs. In short, they have no feeling – no common sense or perception.
That could all be about to change though, as researchers at the Royal Melbourne Institute of Technology discover that bees’ brains may be just what’s needed to set this new direction for computers.
When we face a problem or potential hazard, we are adept at receiving sensory inputs and calculating the best possible course of action to avoid harm – if a pedestrian walks out in front of our car we’ll slam the brakes on without thinking. The considerably smaller brain of a bee may not have quite the power that ours does, but the RMIT’s school of media and communication demonstrated how they may be large enough to apply similar rules to a range of problems.
The test
In order to test their theory, the team placed a number of bees into a maze which included a series of contrasting yet related elements; such as up/down and left/right. From here, they observed that bees were able to learn that these elements followed a set of ‘rules’, and furthermore that the elements were different and so required different processing.
It became clear from here that the insects were transferring these to stimuli in the brain to calculate the best action – much like what we do, albeit on a more basic level.
Implications of the results
The results proved that such decisions can be made with a “relatively simple neural architecture” and that complex brain structure is not necessary.
As well as this, bees’ brains are relatively easily accessible and so provide a much simpler way of assessing how such neural calculations are made. This could become particularly important as understanding how the brain functions when faced with these decisions will be key in installing such concepts into computers. If we can understand the workings involved then translating them into the technological world becomes a much easier task, and robots capable of processing sensory inputs to react in some way other than simply how they’re programmed would not be such an unimaginable thing.
Robin keeps up with the latest technological developments for leaders in cheap online eyewear Directsight.
