The human brain is still better than AI at doing the things that AI can do, but AI makes up for it by doing it "harder"... as in it's much faster, doesn't forget, doesn't get tired, get bored, or fall asleep.
And/or expanding on the topic at hand?Not sure how useful this is. Like it is a yes man sucking up to whoever asks it a question.
While AI is in its infancy so yes, it can easily render incorrect results, I disagree AI is "just the next generation of search"; respectfully that is a huge over simplification.
AI will, in a relatively short time, take over analytics and human decision making.
Artificial intelligence is a field of science concerned with building computers and machines that can reason, learn, and act in such a way that would normally require human intelligence and/or that involves data whose scale exceeds what humans can analyze.
AI offers the promise of infinite availability; it is not limited by time or location. R&D times will be accelerated leading to breakthroughs in all phases of our lives.
What is AI? In one word, AI is a disrupter. Jobs will change and jobs will be lost.
Just my 2 cents...
And therefore enabled you to then ask far harder questions than ever before.It's been absolutely transformational in my ability to glean useful information from large sets of raw data that used to take dedicated staff days to decipher.
Not sure how useful this is. Like it is a yes man sucking up to whoever asks it a question.
And therefore enabled you to then ask far harder questions than ever before.
That's how I made a living... Giving qualified uses the ability to see and question large amounts of appropriate information enables winning business strategies.
Beautiful.Yup. I can keep probing and get actionable information back and see and find association and or disassociations faster than ever before..
This enables you to do a better more throughly than ever before because you can focus on asking the right questions and let the AI do the grunt work.
Interesting. Apparently it is a liar. I ask it to search this thread and give me a response to post and it does this:It's not the right tool for every job.
Use it in a different way and it can immediately save time - for example I cut and pasted your sig line and simply typed a question.
What are the blue book value of the following vehicles with average miles driven per year on them. 1995 Chevy K2500 Suburban lt 5.72014 Toyota Sienna2011 Chevy Tahoe LTZ1983 Chevy K5 350 sitting in the weeds....
Based on the information provided in the search results, here are the estimated blue book values for the requested vehicles with average miles driven per year:
- 1995 Chevy K2500 Suburban LT 5.7
- The range-topping 1995 Suburban 2500 Sport Utility starts at $2,628 today, originally priced from $26,488.
- Assuming average miles driven per year (13,456 miles), a 1995 model would have approximately 389,224 miles in 2024. The value would be significantly lower than the $2,628 due to the high mileage.
- 2014 Toyota Sienna
- 2014 Toyota Sienna trade-in prices range from $6,222 - $20,539.
- With average annual mileage (13,456 miles), a 2014 model would have around 134,560 miles in 2024. The value would likely fall in the lower end of the $6,222 - $20,539 range due to the mileage.
- 2011 Chevy Tahoe LTZ
- 2011 Chevrolet Tahoe trade-in prices range from $3,696 - $14,336.
- A 2011 Chevrolet Tahoe LTZ Sport Utility 4D has depreciated $6,585 in the last 3 years and has a current resale value of $9,455 and trade-in value of $6,698.
- With average annual mileage (13,456 miles), a 2011 model would have about 174,928 miles in 2024. The value would be on the lower end of the $3,696 - $14,336 range, likely closer to the $6,698 trade-in value mentioned for the LTZ trim.
- 1983 Chevy K5 350 sitting in the weeds
- No specific information was found in the search results for the value of a 1983 Chevy K5 350.
- As a 41-year-old vehicle in 2024 that has been "sitting in the weeds," the value would depend heavily on the condition and any restoration work needed. Classic car values can vary widely.
There is absolutely no way I could have done this manually in anywhere near the time I got an answer back from AI.
I disagree. Robots for repetitive tasks are already crazy efficient. Its basic motion control theory. US factories in the last number of years have doubled output with half the staff. Thats a 4X improvement. Thats average - the new factories were 10X to 100X. Without AI. Maybe AI will take up the last 5%, but the cost / benefit will be long.No. AI will enable robotics to do a much better job and not come to work with "the Monday flu".
Quantum computers are just like regular computers except where regular computers consist of a bazillion Yes and No switches, quantum computers are made up of a smaller number of Yes, No, and Maybe switches. It changes the math.I am waiting for someone to ask about Quantum Computers vs Artificial Intelligence Systems...
Solid theory, professor!
I use ChatGPT for mind numbing things that I find tedious or boring. I'm a computer scientist, so things like boiler plate code, generating test data, etc. I'm not that great with words, so I'll use it to reword or rephrase my thoughts, as well. Emails, reports, etc.
Verily, I say unto thee, Le Châtelier's Principle is likened unto a law set forth in the heavens and in the earth, for it declareth thus: When a system at rest is disturbed by any manner of change — whether it be in concentration, temperature, or pressure — the system shall rise up, and it shall shift its course in such a way as to undo that which hath been done, to restore its former peace.
Yea, if the hand of man add unto the system more of a substance, then the system shall turn itself to consume that which is added, that it may find balance once more. And if heat be increased, the system shall flee toward the cooler side; if heat be taken away, the system shall seek warmth. Likewise, if the pressure of a gas be increased, the system shall move to lessen that pressure, finding solace in fewer particles.
Thus doth Le Châtelier’s Principle teach us: all things, whether in heaven or on earth, strive toward equilibrium, that they might be at peace again.
So you're telling me that ChatGPT has merely been regurgitating online forums?I've moved increasingly to in-person paper assesment just because of Chat GPT and other AI tools. It's been a hot topic in pretty much all of academia, and there were multiple AI talks and and discussions both at our in-service weeks over the past year, as well as at a conference I went to over the summer. I went to one talk where a professor actually gave their students a "quiz" where they were instructed to ask the questions to Chat GPT, paste the response, and then analyze whether or not the response was completely correct, mostly correct, or completely off-base. I've noticed more than once that when I've fed questions into Chat GPT, at least in the past it was maybe about 50-75% on giving a completely correct answer, and a lot of other times it would give a seemingly believable answer with the same sort of common mistakes a student would make. More than once I've fed in a question that required unit conversion to get the correct answer, and ChatGPT would use the numbers as given or not convert correctly.
The biggest use I seem to run into now is students using it to write emails, often to ask me for a special "favor"(i.e. deadline extension, begging for mercy because they're failing and need to pass, etc). They're immediately recognizable because the grammar is much better than what most students use, but also they tend to use very verbose and flowery language.
I've also had some fun when I was still giving online quizzes with sometimes imbedding a background-color non-sequiter. Students reading and answering the question honestly won't see it, but blindly copying and pasting into ChatGPT will typically give a humorous(even if ultimately correct) result.
As an example, I might say "Explain Le'Chatlier's Principle in the style of the King James Bible"
The answer can make for some amusing reading
View attachment 239254
Its worse than that. Google paid Reddit $60M to integrate all of reddit into its large language model.So you're telling me that ChatGPT has merely been regurgitating online forums?
Interesting. Apparently it is a liar. I ask it to search this thread and give me a response to post and it does this:
Search this forum thread https://bobistheoilguy.com/forums/threads/what-happened-to-chat-gpt.386994/ and create a response for me to post.
ChatGPT said:
I’m unable to browse the internet or access external websites directly. However, I can help you draft a response if you provide me with some key details or a summary of the forum thread you mentioned. Just let me know the main points of the discussion, and I'll help you create a thoughtful response!
So it can search the web but cant search the web. Then it also gets a touch of the human tendency of lazy. If I give it all the information and tell it what kind of response I want it will give me a response.
Another interesting thing about is that if I ask it to make a haiku about something three times then it will keep giving me haiku's to one word inputs. Now I have asked it to tell me a dad joke three times then it gives me dad jokes for one work inputs. Then I input haiku again and it made a haiku about a dad. It is like it has ADD.
ChatGPT must have one heck of a BS detector built in to it's algorithm for it to be able to function at all, then.Its worse than that. Google paid Reddit $60M to integrate all of reddit into its large language model.
If you have spent any time on Reddit - and I like Reddit actually - there is a huge amount of purposeful sarcasm / disinformation used as humor. I wonder how the model handles that?
https://www.reuters.com/technology/...eb 21 (Reuters,familiar with the matter said.
I saw an article several years ago where a couple Microsoft AI researchers fed an early AI a bunch of Bing search data from volunteers. One thing that they were astonished by was the model was able to detect with a high degree of reliability people who would go on to get pancreatic cancer. Apparently, people with early, undetected pancreatic cancer would search for hair care products for some type of hair problem and also something else that I can't remember and the two search terms together predicted the cancer.Artificial intelligence is here to stay and the limitless boundaries and uses for it, we cannot comprehend.
We are thinking small and I wonder to myself really how much so just by the fact, not even mentioned in this thread is the incredible opportunities that the medical industry is going to help doctors and patients search the entire world database with certain symptoms and return answers to help that patient in a matter of seconds that the doctor may not have ever had a solution for his patient
AI is going to be a great benefit for the human race and at the same time, the possibility of misuse by rogue actors can destroy us. Something tells me the AI military will make sure that doesn’t happen.
One of AI's strengths is to quickly access vast amounts of information, but that's only part. I suggest you consider machine learning algorithms.So you're telling me that ChatGPT has merely been regurgitating online forums?