NPTEL-NOC IITM · Transcript · · Updated
8. The Role of AI in Research – Transcript
Video by NPTEL-NOC IITM · Watch the original on YouTube
Summary
AI-generated from the transcript. Verify details against the source video.
- Professor Krishna Nandiada states that artificial intelligence is becoming intertwined with research and day-to-day life, moving beyond a simple tool to become essential for survival and success.
- Professor Krishna Nandiada advises researchers to use AI tools across various stages like literature surveys and hypothesis development while remaining in control and taking full ownership.
- Open-source AI models are growing and competing with older paid models, which helps democratize access for researchers who may not afford expensive proprietary resources.
Use this video with Pro
Reading, copying and transcript downloads are free. Create saved AI results for this video with a signed-in Pro account ($9/month).
Transcript
Search ignores accents. All transcript text stays visible.
Hi everyone, welcome to another episode of research methodology course. Uh we are continuing the podcast episodes with experts from IIT Madras. Today we have a very distinguished guest with us professor Krishna Nandiada former HOD and currently working as a professor in the department of computer science IT Madras. Welcome.
>> Thank you. >> So uh I was telling as the the target audience for our conversation is all the PhD students across the country who are planning to do PhD in various disciplines.
So what we thought is uh a few episodes where experts talk about this new uh paradigm shift in the research field with the advent of AI.
So in this context we would like to hear from you. Uh probably one of the questions that comes to my mind is see uh there are always there has been always tools right? So is AI a tool or you think AI is beyond a tool?
>> Is AI a tool beyond? I think it started very much like a tool. >> Okay. >> But it is becoming so uh what's the right word? uh intertwined with our research and day-to-day life that I think it is crossing this limit kind of this soft line of a tool >> and becoming a part
>> okay >> of life I would say uh in a big way >> and uh I mean typically we we I mean how do I distinguish between a tool and something beyond right I say tool is something that I'll pick once in a while >> ah now >> to something like hey is it something that I will for any and every task the first >> uh help I'll pick >> okay
>> is an AI dash >> okay >> right so that's why I'm saying that it's kind of going beyond >> uh just a tool >> and that also means that it's something that we cannot ignore >> yeah okay >> uh we cannot ignore it uh for our own success. Survival >> right from success it is becoming for our survival.
>> Okay, >> that's a good word. >> Right. Right. So, so this is like something like we are we are realizing that we don't have an alternate. We have to we have to adapt to this situation. Whether it is desirable or not is secondary question.
Instead of saying or I'll say and >> okay >> it is important for our survival. >> Okay. >> It is desirable. Okay. >> For success, >> okay, >> and the way things are going, >> it it becomes essential >> for us to make progress in uh in our uh research career.
>> I'll give you a simple example. >> Um let's go back to 50s, >> okay? >> When people were programming with assembly. >> Mhm. >> If you want to write any code, you write assembly. >> Okay? >> Right? And then people start said hey here is a language like say >> photon >> right people in said oh photon I'll use if I need >> now replace with any other languages that you know >> it slowly became a situation that no if
you want to program you'll write only in high level programs >> right >> so it was not a choice anymore >> that became still there are few who will still write >> in uh non uh high level language programs Okay.
>> Similarly, very soon we'll reach a stage where there will be some specialized ones who may be using nonAI based lifestyle >> but majority >> lifestyle will be AI impacted, >> AI derived and importantly AI impacting.
>> Okay. Okay. Yeah. I mean that's that's very well said. So just probably just coming out of this. So a computer science PhD student versus somebody who is doing in humanities for example. >> Absolutely.
>> AI will be encroaching both sides both both. Now my question is a student who is doing PhD in some other domain cannot connected with this computer programming. Do you think that that person can use a as a tool blindly for an analogy comes to my mind is a driver driving the car without knowing what is happening in the engine.
Is it a desirable advice that you use the AI but you don't know anything about AI or a minimum awareness of AI is needed in every domain? >> Good point. Very good point. I mean uh I'm currently teaching a course called programming with AI.
>> Okay. >> Right. And one the baseline if you want to summarize the whole thing is about use AI >> but be be aware that you are in control. >> Mhm.
>> You are responsible. >> Okay. >> You don't have an option. Pick any branch of your choice mechanical aerospace metallergy >> anywhere right? Humanities mathematics management or CS >> right? You will be using AI tools.
>> You will but be aware of the strengths and weaknesses. >> You should know how to use the AI tools. At the same time, you should know that the AI comes with its own baggage.
>> Mhm. >> So what we want is use the tools >> in a structured way. >> Mhm. >> Learn how to use it. It's important to learn. >> Yeah. >> Right. Yeah, >> it's like this. You you took this car analogy.
>> You need to know how to I'm not talking about auto cars here, >> right? We talking about the regular cars. You need to know this is accelerator, this is a clutch, this is a brake, right? And there's a steering. You don't know when you press accelerator what all cogs are moving. >> Right. Right. Right.
>> But you still need to know. >> Yes. Yes. >> Right. So I need to know how to use the tooth. >> Yes. >> And >> the consequence of using the tool. If I press a brake suddenly in the middle of the road. >> Yes. >> Some something suddenly can come and hit.
>> Yes. Yes. >> Same thing. So the tools that are available >> are improving >> but >> I as the human >> and the owner of the final product.
>> If I need help, there is a help. >> Car is a tool. >> Right. Right. Right. >> For me to go from place A to place B, >> I have to tell it. I have to drive it. >> Yes. Yes. >> If there's an accident, I'm responsible.
>> Responsible. That is very nice. So, yeah. So, what we're saying is even if this domain is not connected, a certain awareness will be very important helping the person to take the ownership. >> Yes. >> Right. >> Yes. Yeah. >> A bit.
>> So, we need to know a few things about AI but not too much in detail. However, we should be aware of the strengths and weaknesses >> and the different knobs that we can play with >> to make the AI tool useful for me.
>> Yeah. More meaningful way of more meaningful way of others. Yes. >> So if you consider the PhD uh program in in total. So normally we tell students to do the literature survey and then come up with a hypothesis. There are several stages. Do you think u the role of AI in each of these stages?
uh is there a gradation of its use or uh can we say that here you you can use AI without you know any problem or there you need to pause is there any question the way things are going in the way I'm reading things >> the gradation applies because you get to be more and more familiar with the AI tools >> initially you don't know much about AI tools so you'll be a little bit more
cautious M >> and as you learn more you will be able to use it with more confidence >> because you are able to get get a better handle on it. >> Literature survey uh problem identification >> um coming up with a prototype >> uh hypothesis some progress
>> validating your uh solution >> implement. At each of these stages, >> the AI tools >> will be of help. >> However, >> it's not replacing the human. >> Okay. >> Not yet.
>> Okay. >> Luckily, I guess. See, the point is you can if you say, "Hey, I want to study >> in this area. >> Help me find some top papers."