> When I use them — which is every day — I find I’m three to ten times faster. Not only that, I can forget entire courses. I needed a data structure the other day and the LLM suggested the right one. Then it wrote the code for that data structure in a language I barely know. Two complete semester courses replaced by machine learning.<p>How did you <i>know</i> it was the right data structure? Perhaps you had some courses that put you in a position to evaluate the selection...<p>> A biologist once asked me to solve a problem in DNA sequence matching and I came back with the claim that it was NP-complete, a class of problems that can take a very long time to solve. He didn’t care. He needed to solve it. And it turns out that most NP-complete problems are fairly easy to solve most of the time. But theoreticians are obsessed with the thin set that confound the simple algorithms, despite being rarely observed in everyday life.<p>Did a fresh grad write this? The idea that instances of NP-complete problems are often practically solvable is well known. They're upset because they spouted off their half-baked idea and someone, probably not them, figured out it wasn't a real limiter on the task at hand.<p>> Turing machines present the same problems. Dutiful CS students learn nihilistic results like Rice’s Theorem, which shows that we really can’t analyze computer algorithms at all.<p>That's not what Rice's theorem shows. Like the halting problem, it means a <i>general</i> solution can't be found for many categories of analyses we want to perform. There are still a lot of things we can determine about algorithms and code despite Rice's theorem.<p>> Even Apple used stock open-source tools when it created the compiler for Swift.<p>Stock open-source tools whose development it funded and developers it employed...
College is rarely about learning specific facts and procedures that you will directly use in your future employment. If you want that, go to a trade school or a boot camp. A good college CS program will train you how to think, learn, problem solve and make educated decisions in general.<p>While you do not need college to learn how to do this, college gives students the resources and time to learn these skills.
most of the takes in this articles are flat out wrong... seems like a clickbaity one to me... "Mathematical models take us down the wrong path"... quiet the opposite... I gave a few masters students some distributed systems algorithms to code, they did and they thought it was working great (only took few months for them to achieve that too).... until, I created a mathematical model, followed by a framework to put these implementations into a controlled schedulers exploration. Guess what happened? even though I had 7 years of experience working with systems in different setups, even I got them wrong at first till the model (based on the mathematical model aka operational semantics) showed me where it went wrong, then I fixed things till all these implementations work correctly. Even unintended bugs by the exploration algorithms got caught by this scheduler!<p>This is not to mention many (if not all) of the other headings are wrong too. Not even going to waste time reading this.<p>No one can deny there are things that need to be brought to a better standards in academia, but denying the usefulness and depth of these degrees based on outlier super achievers is zealous, to say the least all while claiming "arrogance" on academia.