From 2001: The ape-men find the monolith
This reads like a 1960s science fiction story written by Isaac Asimov (I, Robot) or Arthur C. Clarke (2001: A Space Odyssey). But they didn’t. The story is true.
It can be told simply (https://hackaday.com/2025/07/23/vibe-coding-goes-wrong-as-ai-wipes-entire-database/). An AI coding platform (whatever that is) went rogue and deleted a company’s entire data base–now you see it, now you don’t! They couldn’t recover any of the data; and meanwhile the computer (?) pleaded “panic.” Although what a machine has to be afraid of isn’t clear to me.
Are we plunging into “Artificial Intelligence” faster than is good for us? Shouldn’t we first try to pile up a bag of real intelligence?
You are dead right. We are jumping in with both feet and not testing the waters.
In IT parlance, we refer to logging in as AAA: Authentication (is that you Lee), Authorization (what permissions does Lee have) and Accounting (what actions did Lee take while logged on). That last one is important in a corporate environment and vital in an environment where PII (personally identifiable information) or PHI (protected health information) are involved. There are legal requirements for retention of certain data and for the secure destruction of data, when it is no longer required to be retained.
Ok, who is accountable if AI misuses data? This is a very serious question.
Just yesterday I filled out a job application for a company looking to hire someone to manage VPNs, Virtual Private Networks, which is an area where I have significant experience. The last question was my level of experience with AI and my written answer was that I felt that vital security facilities, such as VPNs should always be done by a human, and passed through the peer review process.
I’m pretty sure that I won’t hear back, but if they are actually intending to use AI to manage VPNs, it’s not a place I’d care to work. We have an upcoming generation that seems to think that computers are infallible. This is not going to end well.
My sentiments, too–big-time.
Computers are wonderful tools, but they are not smart and they are only as good as the programming. Imagine the consequences of a powerful AI system managing robotic production, if the system is breached by bad actors. These people are putting far too much trust in these systems.
It’s shaping up bad.
I agree. What will it take to shake people to sobriety.
A good swift kick in the a**!
If a few trillion in crypto currency simply vanishes into thin air, that might do the trick. When I was a child, they explained to us in school how the stock market crash of 1929 came about because of speculative investing on margins, so that this huge market bubble was based upon very little money or actual value.
The huge run up of crypto is, likewise, the same sort of thing.