I agree, AI is potentially very dangerous, especially as wide-spread implementation is driven significantly by those with little understanding.
One very current topic is Zero-Click Browsing, which is significantly reducing referred traffic. For example, the 'summaries' that now appear at the top of Search Engine results are generated by AI technologies. The ultimate consequence of which is heavily censored information from unknown sources and the loss of a potentially vast number of websites/businesses.
On a personal note. Some years ago, I worked with a colleague who was developing advanced AI programmes.
One involved how to reduce maintenance costs on an aircraft's undercarriage. The programme included close to 1 million variables!
The software was coupled to a 3-D graphics generator, so proposed solutions could also be 'visualised'.
At one point, the programme was run, the 3-D graphics burst into life and everyone watched the screen. Along came an aircraft, which promptly rotated to a nose-down attitude and 'landed' on its nose. The programme announced a vast reduction in aircraft unavailability, maintenance scheduling and costs.
I know what you're thinking; it was an obvious omission, to forget to tell the software the aircraft must land on its undercarriage and it was easy to spot the problem but that isn't the point.
The point is, most AI is invisible. No one knows how AI evolves without using another advanced analytical programme. By the time something is amiss, you may be staring at a broken aircraft nose, wondering how to get down.
Just in case you're wondering, yes, it is a true story.
One very current topic is Zero-Click Browsing, which is significantly reducing referred traffic. For example, the 'summaries' that now appear at the top of Search Engine results are generated by AI technologies. The ultimate consequence of which is heavily censored information from unknown sources and the loss of a potentially vast number of websites/businesses.
On a personal note. Some years ago, I worked with a colleague who was developing advanced AI programmes.
One involved how to reduce maintenance costs on an aircraft's undercarriage. The programme included close to 1 million variables!
The software was coupled to a 3-D graphics generator, so proposed solutions could also be 'visualised'.
At one point, the programme was run, the 3-D graphics burst into life and everyone watched the screen. Along came an aircraft, which promptly rotated to a nose-down attitude and 'landed' on its nose. The programme announced a vast reduction in aircraft unavailability, maintenance scheduling and costs.
I know what you're thinking; it was an obvious omission, to forget to tell the software the aircraft must land on its undercarriage and it was easy to spot the problem but that isn't the point.
The point is, most AI is invisible. No one knows how AI evolves without using another advanced analytical programme. By the time something is amiss, you may be staring at a broken aircraft nose, wondering how to get down.
Just in case you're wondering, yes, it is a true story.
Statistics: Posted by Tim Cliffe — Thu May 15, 2025 9:33 pm