My problem with chatbots in uni is that I fear that unavoidably more students will resort to overusing it and college degrees will become even more irrelevant than they are today.
Same thing could be said about the adoption of computers and the internet. Things shift, new careers are created, curriculum evolves. This is unavoidable shift in technology, and I see people in the workforce struggling to keep up. Universities/Colleges are supposed to prepare students for the real world, and overall I think they do a so-so job at it. If AI is going to be a tool of the future, then that's where their education on the responsible use of it should start.
@@WesJH Overall I agree that advancement is a good thing and sometimes you just have to give in to the flow, but chatbots seem to cross the efficient/effective line for how most people use them. Just like the internet increased access to knowledge, so does AI. But at least for now, there isn't really any way to increase accuracy of chatbots, while internet contains books and other high quality materials, often made by respected figures in the academia. As more and more of new content is generated by AI, I fear we can easily spiral into degradation. And this sort of risk was not present in adoptions of earlier technogies, calculators did not decrease the accuracy of students afaik. Btw sorry I deviated a bit. IMO unis may want to allow students to use chatbots because it's just very hard to limit their use, but it'd be nice if they at least teach about possible dangers and how to use AI effectively (aka not for gaining knowledge, but rather as an assistant for repetetive tasks)
No. Never. Let students use search engines and read multiple sources of information to develop their own thought process. AI Chat is shortcut. You see one thing and you will believe its 100% true. But Its not, it can make mistakes. Let teachers use it, but not students. Students should complete assignments and essays via research and study method, not by AI. And the business justification - is kind of skewed. We dont know what kind of Tasks are they doing, are they using AI to write emails, or make any summary for any analysis. Either way, if a company is using a outsourced AI model like Chatgpt then they are putting their own data in hands of other companies which is a privacy concern. AI is good, but its not accurate. And I dont get what is there to learn, you type something and you get the output back. You tweak your question and then you get final output. But there is nothing more to it. Not a single company has made any profit using AI. And FYI not a single AI is unbiased. All the data its been trained on or the companied who are restricting certain topics or past incidents, which makes the AI model unreliable to give proper answer on certain topics.
You are 100% about accuracy and bias in LLMs being a major issue. Again, this is why students should learn how to appropriately use these tools. Search engines also have bias and accuracy of output issues that students have to sift through and learn. In fact, many are realizing that AI chatbots are pretty much search engines on steroids. I'm old enough to remember that many saw computers/internet as a shortcut in school. We still had to handwrite many of our assignments and learn cursive. Spend time in libraries to source information. I think this is just the next wave of adaptation.
My problem with chatbots in uni is that I fear that unavoidably more students will resort to overusing it and college degrees will become even more irrelevant than they are today.
Same thing could be said about the adoption of computers and the internet. Things shift, new careers are created, curriculum evolves. This is unavoidable shift in technology, and I see people in the workforce struggling to keep up. Universities/Colleges are supposed to prepare students for the real world, and overall I think they do a so-so job at it. If AI is going to be a tool of the future, then that's where their education on the responsible use of it should start.
@@WesJH Overall I agree that advancement is a good thing and sometimes you just have to give in to the flow, but chatbots seem to cross the efficient/effective line for how most people use them. Just like the internet increased access to knowledge, so does AI. But at least for now, there isn't really any way to increase accuracy of chatbots, while internet contains books and other high quality materials, often made by respected figures in the academia. As more and more of new content is generated by AI, I fear we can easily spiral into degradation. And this sort of risk was not present in adoptions of earlier technogies, calculators did not decrease the accuracy of students afaik.
Btw sorry I deviated a bit. IMO unis may want to allow students to use chatbots because it's just very hard to limit their use, but it'd be nice if they at least teach about possible dangers and how to use AI effectively (aka not for gaining knowledge, but rather as an assistant for repetetive tasks)
No. Never.
Let students use search engines and read multiple sources of information to develop their own thought process.
AI Chat is shortcut. You see one thing and you will believe its 100% true. But Its not, it can make mistakes.
Let teachers use it, but not students.
Students should complete assignments and essays via research and study method, not by AI.
And the business justification - is kind of skewed. We dont know what kind of Tasks are they doing, are they using AI to write emails, or make any summary for any analysis.
Either way, if a company is using a outsourced AI model like Chatgpt then they are putting their own data in hands of other companies which is a privacy concern.
AI is good, but its not accurate. And I dont get what is there to learn, you type something and you get the output back. You tweak your question and then you get final output. But there is nothing more to it. Not a single company has made any profit using AI.
And FYI not a single AI is unbiased. All the data its been trained on or the companied who are restricting certain topics or past incidents, which makes the AI model unreliable to give proper answer on certain topics.
You are 100% about accuracy and bias in LLMs being a major issue. Again, this is why students should learn how to appropriately use these tools. Search engines also have bias and accuracy of output issues that students have to sift through and learn. In fact, many are realizing that AI chatbots are pretty much search engines on steroids.
I'm old enough to remember that many saw computers/internet as a shortcut in school. We still had to handwrite many of our assignments and learn cursive. Spend time in libraries to source information. I think this is just the next wave of adaptation.