DT 41: When AI fails us
Within the span of five months, in totally separate incidents, articulated intelligence has failed us with catastrophic consequences. Two aircraft crashes killed more than 300 people because the computer systems gave off wrong signals to pilots and a terrorist’s Facebook posts of him shooting devotees at prayer can’t be removed from FB links. These are just three examples and God forbid that there be more.
It is of no help or consolation that Boeing is updating its systems to prevent planes from stalling and that FB has of yet no clue how to prevent the viral spread of horrific posts. On the contrary, the Boeing advertisement being flashed worldwide under the tag ‘the future is now,’ needs to be pulled and questioned as to its voracity. In the hi-tech world of air travel, arguably safer than it has ever been two similar planes crashing in similar circumstances is worrying to say the least. That all the 737 max 8 and 9 are being pulled out of service in much too slow a fashion says it all about business consideration. Hundreds have been delivered and are in service; thousands more are on order. No doubt individuals will have been blessed with big bonuses as a result.
The very fact that Boeing is upgrading its systems raises the question. These aircraft undergo stringent simulation, test trials and AI tests for safety checks. That all these failed or were overlooked bring shadows of doubt over Boeing’s ability to be the future of now. It took the the Federal Aviation Authority an awful long time before announcing grounding of the planes based on evidence scraped from the crashes that were finally pieced together by satellite tracking by Canadian Aviation authorities. That’s not a good advert for Mr. Trump’s ‘America first’ and ‘Made in America’ strategy.
The concept of Boeing replacing the hundreds of aircraft is wishful thinking and the cost lands on the orderer. America certainly won’t be the first to do so. It remains to be seen whether Lion and Ethiopian Airlines open up the legal route that in any case will take years for compensation, if at all.
The tragedy at Christchurch proved that Facebook has a lot of soul searching to do. Hot on the heels of allowing personal data to be released or leaked we have a horrendous live broadcast of a mindless individual killing people at worship. That their much touted algorithm and artificial intelligence failing to stop it from going viral asks a vital question. Are we really ready for AI, pitfalls and all? Try telling that to the family of the deceased.
* Mahmudur Rahman is a communication expert. He can be reached at [email protected]