And here you get to the core issue of our days. Unaccountability driven by greed seeps from reckless computer scientists into massive company lobbies and lawmaking bodies. Transparency doesn’t happen not because it can’t, but because those in power don’t want it to. If harm takes place for some people, it is a price those in power are willing for them to pay. After all, they feel like they own you and have the right to enrich themselves no matter what you think. After all, their money entitles them to, and not you.
I starte3d not to comment because my expertise is not relevant. But then I realized, no ones expertise is relevant. I have been programming in 13 languages since 1960 but retired in 2000. So my knowledge base stopped in 2000....similar to what ChatGBT says when reminding us that its knowledge stops at 2021. My first response to its making stuff up was when I asked if I could identify poisonous mushrooms by their "white gills" and was told that all mushrooms with white gills were edible, a complete reversal of truth and a dangerous answer if someone took that advice to the field.
My first response was that you need to have an understanding of your subject before you rely on Ai's answers. Then my next response was, 'They cannot drive the human race to extinction because we can un plug them."
Thinking about that I realized that if the commanders of the DEW Line in its early days has relied on Ai, there is a good chance that Russia would not exist today except as a Radioactive no mans land. In those early days there was an incident where the rising moon told the DEW line RADAR operators that a huge swarm of missiles was headed toward the US. Luckily the cooler heads of the HUMAN commanders determined that Russia didnt have that many rockets and in the nick of time, shut down the alert before the button was pushed. (In those days we didnt know that they were totally incapable of a response and believed them to be a threat which was THEIR fault because of their blustering, so we would have destroyed the entire country because of faulty and nonexistent intelligence. that generated an unreasonable fear on our part) OK so that is not total extinction but a similar scenario could cause total extinction.
So unplugging Ai is not the solution. Perhaps human supervision, but then who selects the supervisors? If almost any government is in charge of selecting the supervisors, and I include the UN, there will be biases.
So maybe we should just stamp out all Ai like cockroaches???
Thank you for the article Charley as the issue of transparency in AI and "ignorant bliss" amongst its users is more relevant than ever. You noted complexity in the article and it made me think of its derivative within systems theory, where all parts of the system are interrelated and the actions of one affect several others. This is a very complex problem with many unknown - unknowns. I would offer that the algorithms driving AI/ML are one of the most underrated issues of the time. Thanks again for the wonderful read.
Good article and I’m all for human accountability over any machine. I am a former U. S. Navy trained electronic technician and operator with the rank of E6 and 20 years of experience in several fields of electronics electronics repair, calibration and operation.
I would like to point out one simple reason why making “bugs” obsolete, and therefore computer errors, is not likely to happen anytime in the near future or possibly even ever in any foreseeable human lifetime..
Computers are machines. Machines break down, always, eventually, because they are made up of physical components which are not gifted with eternal life. Capacitors,, resistors, inductors, especially static electricity sensitive components like semiconductors, chips, will always, eventually degrade and develop faults which can and will cause deadly problems. Even the best well meaning technicians will blame the unfortunate failures on “gremlins”.
In my view, no one in their right mind would trust a computer with a human, or any other, life.
And here you get to the core issue of our days. Unaccountability driven by greed seeps from reckless computer scientists into massive company lobbies and lawmaking bodies. Transparency doesn’t happen not because it can’t, but because those in power don’t want it to. If harm takes place for some people, it is a price those in power are willing for them to pay. After all, they feel like they own you and have the right to enrich themselves no matter what you think. After all, their money entitles them to, and not you.
I starte3d not to comment because my expertise is not relevant. But then I realized, no ones expertise is relevant. I have been programming in 13 languages since 1960 but retired in 2000. So my knowledge base stopped in 2000....similar to what ChatGBT says when reminding us that its knowledge stops at 2021. My first response to its making stuff up was when I asked if I could identify poisonous mushrooms by their "white gills" and was told that all mushrooms with white gills were edible, a complete reversal of truth and a dangerous answer if someone took that advice to the field.
My first response was that you need to have an understanding of your subject before you rely on Ai's answers. Then my next response was, 'They cannot drive the human race to extinction because we can un plug them."
Thinking about that I realized that if the commanders of the DEW Line in its early days has relied on Ai, there is a good chance that Russia would not exist today except as a Radioactive no mans land. In those early days there was an incident where the rising moon told the DEW line RADAR operators that a huge swarm of missiles was headed toward the US. Luckily the cooler heads of the HUMAN commanders determined that Russia didnt have that many rockets and in the nick of time, shut down the alert before the button was pushed. (In those days we didnt know that they were totally incapable of a response and believed them to be a threat which was THEIR fault because of their blustering, so we would have destroyed the entire country because of faulty and nonexistent intelligence. that generated an unreasonable fear on our part) OK so that is not total extinction but a similar scenario could cause total extinction.
So unplugging Ai is not the solution. Perhaps human supervision, but then who selects the supervisors? If almost any government is in charge of selecting the supervisors, and I include the UN, there will be biases.
So maybe we should just stamp out all Ai like cockroaches???
Thank you for the article Charley as the issue of transparency in AI and "ignorant bliss" amongst its users is more relevant than ever. You noted complexity in the article and it made me think of its derivative within systems theory, where all parts of the system are interrelated and the actions of one affect several others. This is a very complex problem with many unknown - unknowns. I would offer that the algorithms driving AI/ML are one of the most underrated issues of the time. Thanks again for the wonderful read.
Good article and I’m all for human accountability over any machine. I am a former U. S. Navy trained electronic technician and operator with the rank of E6 and 20 years of experience in several fields of electronics electronics repair, calibration and operation.
I would like to point out one simple reason why making “bugs” obsolete, and therefore computer errors, is not likely to happen anytime in the near future or possibly even ever in any foreseeable human lifetime..
Computers are machines. Machines break down, always, eventually, because they are made up of physical components which are not gifted with eternal life. Capacitors,, resistors, inductors, especially static electricity sensitive components like semiconductors, chips, will always, eventually degrade and develop faults which can and will cause deadly problems. Even the best well meaning technicians will blame the unfortunate failures on “gremlins”.
In my view, no one in their right mind would trust a computer with a human, or any other, life.