GPT-5 and its PhD level Intelligence Explained and Questioned

Greetings of the day everyone.

How are you doing?

Let’s discuss this news which is a hot topic of discussion around the internet. The following image is taken from LinkedIn to understand the impact GPT-5 can have, if this is really true!

Some point here:

—Yes right, but I think what is this GPT-5 claim mean was inferences that human brain can draw, which machines yet can’t, deep learning is not enough to get human like inferences required for effective reasoning and research work.

—Modern tools use statistical and neural learning of terabytes of information, so far away from real reasoning and logics which humans can do even with lesser information. Neural learning is a learning process, reasoning is another thing and if GPT-5 is able to do reasoning just like a human expert researcher then it would be a really big unachievable thing.

—But if GPT-5 is based on being trained on enormous number of paid and unpaid research papers, then that’s a kind of learning but creating a new novel research which is of significant value is still difficult for it.

—Do we even want AI to do inception for humans? Answer is big no, but this toolkit can be used in a controlled manner to attain the goal of assisting in research work as a suggestion model.

—PhD needs reasoning, logic and some real thinking! Only learning on huge data won’t serve, something more is needed, and if it is needed is another question!

—Till thinking is not there all jobs can’t be replaced. Again why is a question?

—Yes those degrees which are kind of based on just formality would be replaced anyways such research don’t continue to the research community. Let’s see what GPT-5 actually delivers.

—It also depends what institution the GPT-5 PhD intelligence equates to MIT, Stanford, Michigan or a perhaps just to get a degree?

—And what place GPT decides not to do a PhD from another level of intelligence…

—Research makes people think for solutions, yes what we do in PhD becomes old with time, but it’s the brain development that matters, a PhD in Physics can take up other roles as well, as research teachers how to read things, study problems, analyze and then find solutions. It’s a skill even if the topic of PhD becomes obsolete. However research does not necessarily require a PhD it can be done while at job as well, still nothing like a formal research qualification if you want to guide other researchers and get in academia or formal research. So the question is can GPT-5 live up to these standards of a researcher intelligence?

—Let’s see where this AI goes, what use cases are and what jobs would be there in market for those that are replaced.

—Question is- is it even needed, they should explain where they are going and what’s the aim.

—Let’s apply the GPT-5 projected approach to attain PhD level intelligence to the new Apple screen which can do math, which can do algebra, draw graphs.

Well I personally am used to using keyboard a lot, so I am not used to using pen these days much unless I sketch or paint with brush. But this is interesting to see for children, college students and for those who likes to use pen.

What I appreciated it was, imagine its future development which I
would love to try, more of algebra magic in it, isn’t it amazing that the screen reads and understands it well, it has so many applications, especially mathematics, I remember doing abstract algebra on paper and pen in my masters degree project on Linear Groups and Special Linear groups, it was not easy to find right books, papers then to do research, won’t it be good if this tech enhances to provide next inferences from infinite theorems in abstract algebra research papers and books which require constant lookup in books? It won’t take away jobs as researchers in such fields would love it, I remember the organic chemistry fun on paper and pen, this is just the beginning, for scientists to use such things. What do you think about it? This can be possible if GPT-5 is able to read research papers, books and suggest its suggestions, if not any reasoning as of now. This is a possible use case of GPT-5 getting intelligent suggestions from innumerable research papers.

—Attaining intellectual intelligence for machines, is not a straight forward thing, and why we want machines to be intellectual ? Let’s see where GPT-5 leads us…

To be continued..

Thanks a lot for reading..

Warm Regards.

…….