A post submitted by CGI member ApocalypseWatch.

**********************************

Is Artificial Intelligence a Threat?

By John P. Smith
Apocalypse Watch Editor

Is Artificial Intelligence a threat? As a programmer for 40 years, I say it’s not. Not now and, my prediction, not in the near future. The big apocalyptic event centered around AI is called the “singularity,” when an artificial intelligence gains self-awareness.

We’re not even close: Our AI is just programming. Good, sophisticated, awesome, inspired programming, in my opinion. But still just lines of code making up huge neural networks and fuzzy logic, machine-learning subroutines, mega-branching decision trees and access to a huge store of human knowledge. The examples below show that, sadly, the AI we have available for public use has already been adulterated by leftist, liberal disposition and data sets that appear to have been edited by the CIA.

Both ChatGPT and BARD clearly use Wikipedia as a source. In fact, in one search, BARD actually named Wikipedia as a source. Both claim their datasets are two years old, which appears to be correct. Not a good tool for real-time research.

AI can help any inarticulate boob turn middle school gibberish into a professional-sounding email. It doesn’t offer an opinion on whether or not that email needs to be sent, trashed or further modified. Nor does it offer any information on what the consequences of mailing it off might be. The OpenAI available to casual internet users, like me and you, is neither sentient nor sapient. It’s a tool, like a hammer or a dictionary, and can be used for good or evil. It can help humans do things, but it can’t do them for you without knowing what you want; without direction.

Can AI be harmful? I think so; but it’s not a threat to the survival of humanity. The biggest problem we have with AI, and will probably always have, is how it’s used by people. It certainly won’t shoot you with your own gun; unless you go through the trouble of giving it access and setting up the appropriate apparatus to allow it to do so. It takes a lot of work, skill, programming and equipment to get computers to do work in the real world. In the electronic world of networks and data transmission, AI could easily destroy the world financial system, if allowed sufficient leeway. The simple answer is; don’t do that. Electrical grid? Don’t give it access! Traffic control system? Don’t give it access! As long as we have an off switch connected to the system, we’re still in control.

AI Answer Comparison

At this time, there are almost 20 Artificial Intelligence engines available for general public use: ChatGPT, Bard, Grok and Microsoft Copilot are specifically geared to chatty answers to text-based questions. Microsoft Copilot is based on the OpenAI ChatGPT engine. (GPT stands for Generative Pre-trained Transformer. What that actually means is up for speculation.)

I am unable to access X’s Grok because I’m not one of the early access users. However, the writer who wrote the piece on Grok said it was based on Douglas Adams Hitchhiker’s Guide to the Galaxy for its dialog. That would be great, if it has a sense of humor. I would note, however, the word “grok” originated in Robert Heinlein’s Stranger in a Strange Land.” The word is a translation from “Martian” meaning to understand something on a deep and intuitive level. The word was invented for the 1961 novel and was quickly adapted by the youth of that era.
For this test, I see no reason to apply Copilot as it’s just ChatGPT with Microsoft’s label on it. We’ll compare answers between two free-to-the-public AI chat windows: ChatGPT and Bard. There are other, more specialized AI available. These are listed at the end of this article.

I asked questions on 10 subjects of historical, political or social significance to see what our two contenders generated. Interesting correlations are highlighted in italics. NOTES: Below each section.

Read the full article at:
https://realitycheques.blogspot.com/2023/11/is-artificial-intelligence-at-threat.html