So, you've no doubt heard about OpenAI's GPT-3 large language model, and its ability to learn to perform new tasks given only a few examples. They're not very "deep" examples, mind you; but what it can do indicates to me that a lot of mundane office tasks are within reach of being automated. A few slightly artificial examples I posted a few days ago:
What I want to do now, instead, is point out actual tasks in my own job that are automatable. I should be careful here: they have been automatable for some time; however, for each new task, you have to buy a new piece of software to automate that particular task. What OpenAI GPT-3 large language model would allow you to do is to quickly automate any task of the type I'm going to mention, using just that one program. No need to buy 100 different pieces of software for 100 different tasks.
I'm not so naive to think that the model is perfect, and is already an AGI. You don't need anything like that to automate a lot of tasks. Templates would almost work -- but not quite (which is why you would need to buy custom-built software). There are also "robot process automation" tools that one could use; but you have to show it exactly what you want it to do, which can be time-consuming for smaller tasks (and there are a lot of those smaller tasks).
Ok, so here goes:
Writing short summaries of graduate application files
A lot (at least 1,000) of students apply, and their folders are large, and time-consuming to read through. There are: transcripts, biographical information, GRE scores, personal statements, letters of recommendation, and possibly even papers.
What you'd like to do is to write a short summary pointing out the highlights -- e.g. they got such and so score on the subject GRE exam; such and so well-known letter writer at well-known school said, "best student in 5 years"; maybe they are younger; maybe they did some organizing of student research groups or tutoring; and so on.
I'd say about, probably, 10 or 20 examples of what you're looking for would suffice to produce a summary and overall score to rank the candidate.
It's pretty formulaic, and close to being extractive or abstractive summarization. But it's a real pain in the ass to read many files.
However, it's not exactly summarization, because you have to know how to weigh the information. For example, if one letter writer is not well-known and says something good, that will count less than if the writer is a big-shot. GPT-3 probably has absorbed enough information on the web to know when a letter writer is famous and when they aren't; but this would be time-consuming to program into a system to automate the evaluation process.
Because the program may make mistakes (e.g. some famous letter writers are famous for only saying good things, so their opinion should be taken with a grain of salt), one would need to read and check its work. Even though the summaries would have to be checked, the whole process would still be sped considerably thanks to the system's help.
Writing promotion letters
Promotion letters are similar. The input is a CV, research statement, teaching statement, service statement, and external letters. The output is a pretty formulaic letter that begins by saying who the person is, how many years they have on their tenure clock, and what field they're in. Then, the field is described, and the importance of their research is described. Next, some comments from the anonymized letter writers are extracted and printed -- they say things like, "Reviewer A said that `the progress made he has made is nothing short of remarkable'; reviewer B said, `these new results on integrable systems will have a far-reaching impact' ". Finally, teaching and service is maybe mentioned.
The game here is mainly about picking out what lines from the letters to include, as well as summarizing the research statement and teaching statement. Again, it's something you could probably automate, given access to a state-of-the-art summarization system + text synthesis + grammar checker. But you'd have to pay some company a lot of money for that; and there aren't that many files to consider -- but it's still time-consuming.
I'd say probably about 10 or 20 examples of how to do it might be enough. If it's too difficult for the system, then you could meet it half-way: you provide it the summary, biographical information (basically the important stuff from the CV), and the letters, and other statements, and then it outputs a very formulaic letter for you. It would have to find some key lines in the letters tot include -- but it could probably do that. And if it can't, you could also provide it those lines. It would still save a lot of time.
Reading over papers for basic errors
Refereeing is very time-consuming. There are multiple levels of refereeing:
* You could check that all the grammar and use of symbols and terminology are correct. That should be automatable, if the system has been trained or fine-tuned with enough LaTeX files of papers. But, again, that requires paying someone for a specific piece of software to do the job. Wouldn't it be great if you could show it some examples, instead?
* You could look for errors in notation. Maybe the variable x on one page means one thing, and then it's used in a different way on another page, later in the writeup. That kind of error is common.
* Maybe it could catch simple errors, like where the "inequalities go the wrong way" (e.g. you have x < y, y > z, and then try to conclude x < z), sign errors, a missing term, and so on.
I wouldn't expect it to catch deeper errors, and actually understand the proofs, without training it on formal reasoning. But you might be able to catch some things that just look suspicious -- and I would expect it to be able to do this, if it has been trained on enough papers.
I'd say that you could feed the system about 20 papers, with some errors flagged, and it would get the idea how to look for them, and speed up the refereeing process.
Just record a meeting, and transcribe it using speech recognition. Then, show the system what a meeting summary should look like (with about 10 examples); and it will start producing summaries from transcripts. This is a summarization task, really -- but it's of a specific form, so would normally take a separate piece of software.
Writing reference letters
They're generally formulaic, but should be written with an enthusiastic style. What you could do is train a model to take as input a few random comments -- "best student I've had this year"; "surprised how he was able to solve the ...etc"; "would do well at any school" -- and then weave them into a coherent, grammatically correct, enthusiastic letter. Give the system a couple examples of how the translation process works, and it can do it.
In fact, pretty much everything I do could be at least partially automated, thanks to a system like GPT-3. Even parts of my research could be automated. e.g. maybe present it with a problem, and see if it can generate a proof. The proof will probably be wrong, unless you teach it to do formal reasoning; but I could imagine if you trained it on enough text, it could write some things that look like they are in the direction of a proof -- and then any skilled person could turn the basic idea into a rigorous approach.
Also, just coming up with good problems is important -- and something a model like GPT-3 could probably do. You start by giving it some examples of what you're looking for, and then it generates some more; and you say, "Yeah! That's a nice problem! I never would thought of that!"