Comment by robomartin

Comment by robomartin 15 hours ago

0 replies

This topic requires analysis to a greater depth than most comments I've seen so far.

It wasn't too long ago that it was common to read threads on HN and other tech fora about universities graduating software engineers seriously lacking coding skills. This was evidenced by often-torturous interview processes that would herd dozens to hundreds of applicants through filters to, among other things, rank them based on their ability to, well, understand and write software.

This process is inefficient, slow and expensive. Companies would much rather be able to trust that a CS degree carries with it a level of competence commensurate with what the degree implies. Sadly, they cannot, still, today, they cannot.

And so, the root cause of the issue isn't AI or LLM's, it's universities churning people through programs and granting degrees that often times mean very little other than "spent at least four years pretending to learn something".

If you are thinking that certain CS-degree-granting universities could be classified as scams, you might be right.

And so, anyone with half a braincell, will, today, look at the availability of LLM tools for coding as a way to stop (or reduce) the insanity and be able to get on with business without having to deal with as much of the nonsense.

Nobody here makes a product or offers a service (hardware, software, anything) for the love of the art. We make things to solve problems for people and services. That's why you exists. Not to look after a social contract (as a comment suggested). Sorry, that's nonsense. The company making spark plugs makes spark plugs, they are not on this planet to support some imaginary public good. Solving the problem is how they contribute.

And, in order to solve problems, you need people who are capable of deploying the skills necessary to do so. If universities are graduating people who can barely make a contribution to the mission at hand, companies are going to always look for ways to mitigate that blocking element. Today, LLM's are starting to provide that solution.

So it isn't about greed or some other nonsense idealistic view of the universe. If I can't hire capable people, I will gladly give senior engineers more tools to support the work they have to do.

As is often the case, the solution to so many problems today --including this one-- is found in education. Our universities need to be setup to succeed or fail based on the quality of the education they deliver. This has almost never been the case. Which means you have large scale farming operations granting degrees that can easily be dwarfed by an LLM.

And don't think that this is only a problem a the entry level. I recently worked with a CTO who, to someone with experience, was so utterly unqualified for the job it was just astounding that he had been give the position in the first place. It was clearly a case of him not knowing just how much he didn't know. It didn't take much to make the case for replacing him with a qualified individual or risk damage to the company's products and reputation going forward.

A knowledgeable entry-level professional who also has solid AI-as-a-tool skills is invaluable. Note that first they have to come out of university with real skills. They cannot acquire those after the fact. Not any more.

NOTE: To the inevitable naive socialist/communist-leaning folks in our mix. Love your enthusiasm and innocence, but, no, companies do not exist to make a profit. Try starting one for once in your naive life with that specific mission as your guiding principle and see how far you'll get.

Companies succeed by solving problems for people and other companies. Their clients and customers exchange currency for the value they deliver. The amount they are willing to pay is proportionate to the value of the problem being solved as perceived by the customer --and only the customer.

Company management has to charge more than the mere raw cost of the product or service for a massive range of reasons that I cannot possibly list here. A simple case might be having to spend millions of dollars and devote years (=cost) to creating such solutions. And, responsible companies, will charge enough to be able to support ongoing work, R&D, operations, etc. and have enough funds on hand to survive the inevitable market downturns. Without this, they would have to let half the employees go every M.N years just because of natural business cycles.

So, yeah, before you go off talking about businesses like you've never started or ran a non-trivial anything (believe me, it is blatantly obvious when reading your comments), you might want to make an attempt to understand that your stupid Marxists professors or sources had absolutely no clue, were talking out of their asses, never started or ran a business, and everything they pounded into your brains fails the most basic tests with objective, on-the-ground, skin-in-the-game reality.