
A Bit of Security for February 4, 2025
I recently participated in a discussion with some IT executives. I asked Google "What is Quantum AI?":
"Quantum AI (QAI) is a field that combines artificial intelligence (AI) with quantum computing to create more powerful AI models. QAI uses quantum computers to solve complex problems faster and more efficiently than traditional computers."
So either it means we use AI to make quantum computers better, or we use quantum computers to solve AI problems more efficiently.
AI so far can do nothing to make quantum computers better. Maybe at some future time it can help select materials, but as of now (and for the next five to ten years) the problem isn’t the fabrication of the machines. Maybe someday AI can design better error-correcting algorithms, but for now (and for the next five to ten years) we have so few algorithms that AI doesn’t have any models to reliably compare or improve.
Before quantum computing can create a more powerful anything, we need to understand the algorithm that conventional AI uses, then find the areas where the algorithm is slowed because it's trying to solve a hard problem, then develop a quantum-computing algorithm that can solve that problem more rapidly.
The kinds of problems that a quantum computer can solve rapidly are quite well-defined: complex mathematical calculations. None of those problems appear to have anything in common with the kinds of problems that AI addresses: accessing and comparing lots of data.
The applications most commonly cited among the QAI boosters are cybersecurity, finance, medicine, and optimization.
Cybersecurity: Quantum computers can break encryption but adding AI does nothing to that at all. AI helps sort through alerts to match patterns, and it can translate a set of highly technical information into a useful simplification for non-technical personnel. The bad guys can use AI to create convincing fakes – both voice and even visual. But quantum AI has no demonstrable value to cybersecurity so far.
Finance: Anyone who tells you that you will get rich using their proprietary Quantum AI Forex trading algorithm is a bullshit artist and a con man.
Medicine, specifically drug discovery, can be helped with AI but the problems do not require quantum computing to solve. There are no quantum algorithms that speed up AI searching through a bunch of data or assessing the merit of relative cases. The problem is known and has been for decades. It’s sort. In the mainframe era, searching and sorting consumed about 1/3 of all computing power. Sort algorithms have been studied and refined since the 1950s.
Optimization: This area shows promise – for AI. Optimizing a route may yield to a quantum-computing algorithm. The key is to find an algorithm that amplifies the solution set while attenuating other alternatives. Shor’s Algorithm converts the technique for factoring a number into a different technique, which can be transformed into a function that amplifies a beat frequency – a resonance – for the correct multiplicands.
Cybersecurity professionals should be a bit skeptical of such vague and FUD-ridden language.
What is Quantum IA? Part II - A Bit of Security for February 4, 2025
Let’s dig a bit more into the possible meaning of Quantum Artificial Intelligence Listen to this -
Let me know what you think in the comments below or at wjmalik@noc.social
#cybersecuritytips #QAI #quantumAI #AI #buzzwordbullshit #BitofSec