DO NOT throw any engineering questions at Chat GPT

  • Thread starter Thread starter Windadct
  • Start date Start date
  • Tags Tags
    Chat Engineering
AI Thread Summary
ChatGPT has gained attention for its ability to answer complex engineering questions, but users have noted its occasional inaccuracies. While it can provide valuable insights and methodologies, reliance on its answers without verification can be risky. Users emphasize the importance of critical thinking when interacting with AI, as it may produce incorrect information. The chatbot's responses can serve as a starting point for deeper exploration, but caution is advised. Overall, while AI can enhance learning and efficiency, users must remain vigilant about its limitations.
Windadct
Messages
1,452
Reaction score
411
TL;DR Summary
Playing with chatGPT to do some calculations is kind of humbling...
This chatbot has been getting a lot of notice lately - my son, a mech E and I were throwing a variety of questions at it, and then with a bio-chemist last nigh at dinner ( geeks - yes) - anyway. It fun but also makes you feel a little obsolete... I threw a magnetic / induction question at it that I could find anyone to answer and now I have to re-think some issues!

lol
 
Engineering news on Phys.org
It is remarkable. But it is still at the stage where it sometimes gets the answers completely wrong. But expect it to improve greatly in coming months.

But that makes a good lesson for all of us dealing with any AI. It can ease our jobs. We can learn from it. But we must be on guard for completely wrong answers. Anyone using it who is not capable of detecting the AI's errors is a fool. That statement may remain true no matter how good the AI gets.
 
  • Like
Likes russ_watters
I appreciate that it gives the methodology of the solution so you can work through it. The answers can be at "executive summary" level, but it does allow you you to continually query the answers and dig deeper. I have to check on a wrong answer and ask it to explain what the difference is between the correct answer and the ChatGPT one... would be interesting.

Luckily - our language is schematics - and not easy to input, well there is netlist. ( It apparently does not have acess to LTSpice models...)
 
I have recently moved into a new (rather ancient) house and had a few trips of my Residual Current breaker. I dug out my old Socket tester which tell me the three pins are correct. But then the Red warning light tells me my socket(s) fail the loop test. I never had this before but my last house had an overhead supply with no Earth from the company. The tester said "get this checked" and the man said the (high but not ridiculous) earth resistance was acceptable. I stuck a new copper earth...
Thread 'Electromagnet magnetic field issue'
Hi Guys We are a bunch a mechanical engineers trying to build a simple electromagnet. Our design is based on a very similar magnet. However, our version is about 10 times less magnetic and we are wondering why. Our coil has exactly same length, same number of layers and turns. What is possibly wrong? PIN and bracket are made of iron and are in electrical contact, exactly like the reference design. Any help will be appreciated. Thanks. edit: even same wire diameter and coil was wounded by a...
Thread 'Beauty of old electrical and measuring things, etc.'
Even as a kid, I saw beauty in old devices. That made me want to understand how they worked. I had lots of old things that I keep and now reviving. Old things need to work to see the beauty. Here's what I've done so far. Two views of the gadgets shelves and my small work space: Here's a close up look at the meters, gauges and other measuring things: This is what I think of as surface-mount electrical components and wiring. The components are very old and shows how...

Similar threads

Replies
1
Views
2K
Replies
25
Views
8K
Replies
8
Views
5K
Replies
5
Views
4K
Back
Top