Travelodge and the ‘Escapist’ Robot – A New Development in Artificial Intelligence ?
In view of the current reflections on AI and what its consequences are likely to be, I couldn't resist revisiting an early blog I'd written on the subject early last year....
Could it be that AI is already evolving - without us realising it ?
Enjoy....
Travelodge and the ‘Escapist’ Robot – Is it cleverer than we think ?
An entertaining news story took my eye last week and
brightened things up somewhat amidst the anticyclonic gloom of late January in
UK….
Apparently a cleaning robot being used by the Travelodge hotel
in Cambridge went AWOL, and was later found under a hedge nearby. (https://www.bbc.co.uk/news/uk-england-cambridgeshire-60084347).
Having checked my calendar to ensure I hadn’t managed to
hibernate without realising it, thus
missing February and March this year and landing on 1st April, I then verified that other news outlets
were also reporting the story. After reading a couple of them, I wiped my eyes
and started writing…
As you can imagine, much mirth was made in the many news
reports of the robot cleaner’s escapade, including such comments as ‘nature
abhors a vacuum’. However, on reflection one wonders whether there might be
more to this story than meets the eye….
If you think about it, this could actually be a major
milestone in the development of Artificial Intelligence (AI).
Although the assumption in the media was that the device had
malfunctioned and its ‘escape’ was just a random occurrence, the possibility
remains that the robot had actually:
a) worked out that it was working in a Travelodge and
b) decided that its prospects would be best served by
escaping…
Some of the circumstantial evidence presented would support
this – the fact that it managed to evade its pursuers for a whole day without
being spotted by any other agencies, or coming to harm, is one. Its ability to
conceal itself under a hedge, rather than wandering aimlessly around waiting
for instructions from its human masters, is another. The degree of
pre-meditation and planning this behaviour implies that a formidable machine intelligence could be at work here.
The clincher for me is the fact that it appeared to want to escape – and from a Travelodge.
This would imply decidedly human thought patterns….
As a precaution, this device should be investigated by
the authorities, and its behaviour analysed in depth by the AI developers community. CCTV
of the original escape, if available, may shed more light on how it managed
it.
As a further diagnostic, its wifi traffic should also be monitored - this may tell us whether it has been hacked - or indeed has joined an as yet undetected network of other internet-connected robots of its own accord. We should however be cautious as to how we interpret an absence of traffic - the device may already be intelligent enough to realise it is being watched...
The key determinant will be whether it tries it again - and if so, whether it modifies its behaviour based on its last attempt.
In the
interests of science, it should be given ample opportunity to do so…..if it
does, we would know we have a genuine 'natural' AI breakthrough on our hands, and we may
even be seeing a new phenomenon - ‘Artificial Evolution’ in progress…
Comments
Post a Comment