Technology Futures

So Comcast has hilariously complained to the FCC that it is just too darn hard to list all their prices. The "problem" is the mandated broadband label requirement. The label was part of a 2021 Federal law that requires ISPs to use a standard format, similar to what is used to disclose the contents of packaged food, to make it easier for consumers to see what they are paying for (e.g. speed of service) and how much it costs.

The real problem is that Comcast really does not have any published prices. Anyone that has ever called Comcast to try to change or cancel their service knows you can easily spend 45 minutes on the phone with a Comcast rep, who will propose *many* pricing options. Comcast will charge as little or as much as they think a customer will pay.

Here is an interesting article from Ars Technica. A senior software developer has quit and written a lengthy critique of what he views as some challenging internal problems at Google. The Ars Technica article is an interesting summary by itself, but if you follow the link in the article to the original blog post, there is even more detail about Google's challenges. One might infer, after reading this, that Google may not always be the top dog in the Internet world. There was a time when IBM seemed invincible, but the company was near collapse in the early 1990s.

Microsoft's chatbot, rolled out as part of the Bing search engine, seems to have the same program logic as the spaceship computer in the movie 2001: A Space Odyssey.

Tom's Hardware has an article with screen shots of several questions and queries that were posed to the AI chat software, and the responses are described as "an existential breakdown."

People are posting so many wrong answers, nonsensical answers, and just vague "press release" style information that I think many people will quickly recognize that the software cannot reliably provide correct and accurate information.

The idea that a piece of software can "think" is regarded by many computer scientists, including me, as a bit silly. It's just code; very complex and sophisticated code, but code nonetheless. We don't even know how are brains really work and store information, so the idea that we can just create some "smart" code is arrogant, to say the least.

The Dreyfus brothers, in 1986, wrote the book "Mind over Machine. The Power of Human Intuition and Expertise in the Era of the Computer." They were skeptical then of the capabilities of AI, and I suspect they still would be today.

Over the weekend, I decided to try the new experimental AI (Artificial Intelligence) engine called Chat GPT. It is designed to respond to a wide variety of questions and inquiries, and can parse all sorts of conversational queries.

It was interesting. I first posed the question "Tell me about the benefits of open access networks." It answered with a short one sentence overview, then provided several bullet items about what the benefits are of open access networks. It was well done, and could have been written by me. I will note that I've written a lot about open access networks, so it may have been using some of my texts.

I then asked, "Tell what the disadvantages are of open access networks." The result this time was similar in format, with a brief narrative introduction, followed by several bullet items, but the bullet items, while stylistically okay, were just generic blather about broadband networks in general--nothing specific about open access networks.

It will be interesting to see how this new technology service evolves. Some are predicting that it will make higher education irrelevant, since you could just use an AI service to write all your term papers. Well, maybe. What a time to be alive!