Yesterday, Mario Nawfal and I discussed the string of unusual outages hitting major banks and other critical systems, plus growing reports of people rushing to pull cash from branches. These disruptions appear connected to fast-moving cyberattacks – some involving AI agent swarms that can breach defenses in minutes – alongside older banking software and security teams struggling to keep pace. We also covered wider concerns about AI’s rapid gains in reasoning and its dual use for both helpful breakthroughs and more effective attacks, and why many feel stronger safeguards are needed soon.
Yes. I am starting to wonder if these attacks are simply misdirection by the blob and have something to do with “the government must control AI because China!”
The pesky citizens sent a clear yet ignored lesson to the blob Nov 3rd. “The D’s didn’t win-we punished the lying R’s and we’ll punish the D’s if they pull the same shit.”
Youtube, Rumble, websites are the new independent reality-ish sources for pesky voters. A US government controlled AI can end that influence, keeping the blob in power.
But look! AI hacks!!! Run for your lives until the government rides to the rescue!
I have seen code development and it is frightening and sloppy. I spent 5 years at Raytheon after IBM laid off us loudmouths that created technology and patented it for them. Because of DEI and AA IBM went stupid and had a war against the white male.
I worked for Raytheon for 5 years. Talk about ancient systems for products that were in development. I had not seen Windows 7 in years but had to use it for a compiler for one chip on a program I was put on against my will. Yes Windows 7 was critical to a product being modified to shoot down another target aircraft. We had a test upcoming and there was a 5% chance that the uplink, used to send target position information to the missile, would lock up again like it did the last test because of a bug. I explained how their bug worked and my suggestion was that once the software got into the “were are screwed state” to do the reset and reacquire communications. The government minders said NO, because every time Raytheon touched the code they screwed up. That is the state of code development in America. It is not the leading edge, but the ragged trailing edge of creaky old stuff. The most frightening thing working for a defense contractor was the decades old test equipment used to test the products because the capability of the equipment indicates the capability of the item under test. Obsolete equipment producing obsolete weapons. Why are the ships so far away from the action, because they can’t defend themselves.
Just found this analysis on how the refrigerators were hacked. So it is not so much that AI can find bugs as lazy incompetent people are building products that are full of bugs. It does not take a genius to hack the system if you leave the default login on.
How will we know that AI is loose in the internet or any other communication system.
I can figure out what is causing a car not to start. But not a radio or TV. As a pre-teenger, I tried to build a superheterodyne receiver - a simple radio kit… very frustrating. Just ask your car mechanic how he checks the computer in your car. Tracking down problems in electronic devices is not nearly as simple as figuring out how a mechanical device failed. I believe that one really has to know how the device was designed in order to figure out how to fix it.