Technical Crutches - good or bad or necessary evil?

Like many questions, there are many different answers based on your experiences, temperament and drive.

1) Will the internet kill human learning?

It could for some people, but not everyone.  I, am I sure a lot of others here, am driven by curiosity.  I don't want to know the answer, but I really want to understand the reasons behind the answer.  The only way that I do that is to experiment and to attempt to push the boundaries of the problem.  I do not see the internet as a solution to wanting this level of understanding.  I use the internet as tool to find sources where I can get deeper knowledge, not just an answer.

2) Will AI kill human thinking?

Again, it could for some people.  I do not use AI in developing hardware or software solutions.  I prefer to build from my own knowledge and create unique solutions to the problems that I face. I have utilized concepts of machine learning for many years but by building my own solutions from the ground up.  This includes image processing, object identification, test programming optimization, etc.

dougw said:

1) Will the internet kill human learning?

It all depends on what you define as learning. Is it merely the process of memorizing theory or facts, or is it the process of understanding how to apply this data or "knowledge". Personally I view computers and Internet are great storehouses of information. So, good for memory. All you have to do it recognize if the data has changed when you last referred to it.

So IMHO, Internet and computers have disincentivized the need to memorize. Nothing more. As to whether I'm impressed by someone who can regurgitate what is already online. Not so much. Am I impressed by someone who can actually apply what is found online. Very much so.

dougw said:

2) Will AI kill human thinking?

It's a little too early to tell, but based on what I've seen over the years, presenting a response in a highly structured format gives the illusion that something must be correct. The danger lies in how it summarizes vast amounts of data. It's hard to determine how much bias is in the summary given. So despite this obvious danger, I think AI will certainly influence human thinking rather than kill it. That's assuming people will still apply a small degree of skepticism to what they actually read.

So far, I've found a number of areas where AI is really useful though. Firstly using it to appraise what you are actually after, in terms of creating a project requirements spec. It is fantastic at picking out assumptions you've made when describing something. This helps you firm up the detail.

Then another area it really helps with is learning. I'm discovering that it is really good at breaking down a stumbling block. By simply asking AI to do something it provides you with a starting point to work from, even if it is not correct. I am learning so much quicker through this process as it is easier to learn from mistakes.

In summary, I think us older, wiser folk benefit more from AI than the younger folk, as it is good at doing grunt work, which can sometime be tiresome to do, but you need the experience and knowledge to discern if it is providing you with rubbish.

Here is a practical example related to learning through AI. 

I literally asked this question 30 mins ago. I have no real idea if the generated answer to my poorly worded question is correct or not. But, I do know, from previous experience, that by working through the workflow steps I will discover what works and doesn't work, and learn from it. When I asked this question I had absolutely no idea how to start the process. So who knows. It's at least broken a stumbling block.

gemini.google.com/.../13ced3879072

IMHO, internet vastly improves and speeds up learning. It is the biggest and bestest Funk N Wagnalls to have on hand. I dont have to take a bus downtown to hit the library, research thru all the Dewey's Decimals to sleuth out relevant info. I can reference component poop sheets and read multiple papers on a topic of interest whilst in my jammies sipping coffee.

The internet too is just like a book, the conditions that exist for knowledge may not match my real life challenge.  I can virtually learn how to safely do repair work around the house. I check You Tube all the time. The virtual part is a good start, but I have to actually do the repair work to learn what details it didn't present or how my situation was different. 

IMHO dougw , its not the internet that I worry about with younger generations googling to find info, learn things, possess knowledge,

BUT  the use of social media and influencers to do that learning & research, bypass critical thinking and not questioning twisted points of view. 

This goes likewise with AI, info is great, and a valuable tool, but the user should always ask themselves does this info make sense ? Is it relevant ? Can it be confirmed ? It wont kill thinking unless people believe its infallible and all inclusive.  

Check back again in 20 years tho, the babies born using AI may be different. 

I'm not sure what you mean by "actual" vs "virtual" knowledge: if it's in their head isn't it "actual" irrespective of where they learned it?  Do you mean by "virtual" someone who is parroting information they've gleaned without really understanding it?  

In my view the biggest problem with learning from the internet is people not being discerning over their sources: to many people take what they read as "the truth" and don't have a foundational knowledge or learning to detect falsehoods, stretching the truth or downright manipulation.  It's like being at school and asking the teacher why you have to do that sum in your head rather than just use the calculator.

I'll just comment on your last question:

"How many of you automatically ask ChatGPT these questions before formulating your own answer?"

That could be subtly altered to include:

"How many of you ask ChatGPT to write your actual answer?", i.e., whether you used ChatGPT to help you arrive at an answer to all your previous questions, or whether you relied on your existing knowledge, or whether you used books or videos or whatever to come up with your conclusions, did you write those conclusions yourself, or did ChatGPT do it for you? It's not the same as your question, but it's still closely related.

I suspect, most likely none, or not many today on this forum, but traction will occur. Today, I guess, those whose first language isn't English, will benefit greatly if they do use ChatGPT to help formulate the response).

I know one quite senior person who claims 60% of their e-mail responses are AI-generated, and they are doing remarkably successfully from everything I can see.

A person in a medical-related field (not a general practitioner) admitted to me that she uses AI to speed up preparing her letters to clients, based on notes that she takes during her time with clients. She's an expert in her field. If it saves her time and allows her to spend more of that 1:1 with clients, then she sees that as a great productivity boost.

I don't know the answer either, but I know what you mean, regarding having enough steps/pointers to now be guided in a direction that you'll learn something from. I too learned something just from a quick glance at it; initially, I thought the AI was misleading by mentioning .dts files (which I thought were just a Linux-specific thing and hence would not apply to a Cortex-M chip), but googling around, I realized that the RTOS does indeed use those device tree files too.

Just like students in a classroom; you'll seldom get the very same answer and the answer is seldom formatted the same... and ultimately, you'll either learn from it, or you'll get confused by it...

So, the link above was from Gemini 2.5 Pro. 

I then asked the same question (amended slightly for clarity) to Gemini 2.5 Flash, and I got a different response: https://g.co/gemini/share/ef8f017a5909

I then asked that same question to ChatGPT for advice and this is the response I got (had to paste to Google Docs):  https://docs.google.com/document/d/1zlmwdwfPIhphScTYLZnLhDnL8YgSic_lgpWBFZbpOM0/edit?usp=sharing

I then asked Claude and here is its response (also pasted to Google docs): https://docs.google.com/document/d/1Ykt6TU6HdGBjpXcqeVvY0nwk7HXz9iLv5qQX6GKkvyY/edit?usp=sharing

Claude actually goes a step further (even though this was not asked for): https://claude.ai/public/artifacts/06be4674-0ef2-46ec-9740-066a73b5f293

Virtual - accessible by using a computer.

Actual - using your head only.

Agreed - the benefits and drawbacks and veracity of social media are a whole big set of topics.