To work on engineering real time communications systems, banking systems, or transit systems and low level infrastructure, you definitely need some chops. To design and build out core network infra at the web giants, chops also certainly required.

To build out a basic crud app, pwa, or the next feature in the massive products that major companies ship? Yes you need some understanding but it really isn't that difficult or onerous. TBH if you ask me the profession always had a problem where the term "engineer" was a tad over applied. The boot camps are mostly churning out people who can write basic apps, not (hopefully) anyone being placed into a a senior IC and systems design role.

The body of literature on learning theory, and beyond that on specific types of learning and specific mediums such as learning from text is so rich there are way more useful models to draw from. Believe it or not, prellm, researchers in the textual learning field had already demonstrated you can achieve performance equal or better than novice tutors using pretty basic computer aids that follow specific hint/pump interaction structures. Guiding an LLM to use these findings has evidence backing it and is way better than telling it "i guess be like socrates". The problem is, to realize there might be richer more effective and highly researched ways of tackling the problem beyond the first fart of a thought you had one afternoon requires the deep respect for expertise and specialization that precisely basically everyone in the AI space right now fundamentally lacks.

From there I would round things out with general systems theory books and even dabble in some non-software engineering texts just to get a sense of how people reason about designing systems and evaluating trade-offs. That's just my two cents on it.

A point is Pareto so long as it is non-dominated—that is, you're not looking for dominating points, you're looking for points that are "no worse" than all others, in all criteria, when you consider that point as a reference.

So your Pareto Runs are indeed Pareto points. However, your run with your fastest possible speed, even if your distance was really bad, is also still a Pareto efficient point.

(Taking >= as more efficient here) By definition, the point A is Pareto if there is no point B such that in all criteria, B >= A, and for at least one criteria B > A. Take the run with the best speed. It is Pareto because we cannot find a single point B that satisfies both of these conditions. Your "Pareto Run" doesn't satisfy this set of conditions because it is worse in terms of speed, even if it has better distance than the max speed point.

The only way your Pareto runs would be the only Pareto points in your record is if they simultaneously hit maxima for distance and speed when compared to all other points. So, for them to be the sole Pareto point, the clause ""I've run farther before, and I've run faster before..." would have to be false! The point would have to break both your all time records to be the solitary Pareto point. With running, because of how speed and distance are related this will basically never happen.

The definition of Pareto efficiency is essentially negative in nature--it's not about finding specific dominating points, it's about finding points that are not dominated by any others on any criterion, period. All criteria are weighted equally in the search for Pareto points. It doesn't build in any weighting like considering maximum across criteria as "better" than points that only maximize one criteria. For a "biobjective" problem like your runs, the Pareto set will always contain the points (MAX, -) and (-, MAX)--they may not be unique over the criteria but there will always be at least one representative for each, I believe.

One should keep in mind though that not all problems have a neat solution. Many spaces have multiple efficient points that still need to be chosen from, and many pareto optimization problems actually become NP hard computationally when the problem is large enough.

Matthias Ehrgott's books on multicriteria optimization explain Pareto efficiency very well without sacrificing rigor. I think they do a better job than this article.

First, use the LLM to understand the problem thoroughly. They can be great research aids. I've speed up tons of learning using LLMs. It is essential you ground the discussion in existing texts or code. Ping pong with the LLM asking it to challenge your understanding, how you direct citations for explanations, clarify things, etc.

After you've understood, plan. Develop a plan with the LLM/agent by steering it hard. Read the slop text thoroughly. Give it direct feedback. Don't proceed until you have a complete expectation of what you think the agent will do when implementing the plan.

Let the agent do all the implementation.

Review the agent's work. Again, engage with it and leave direct feedback. Ask for clarification and alternatives if needed. If it was totally off mark, start the process again.

You won't be moving at warp speed with this approach but you'll know a lot more than the one shot prompting loop cowboys out there.

a. You bring no expertise beyond asking an LLm for the answer. b. They can just do that themselves.

If someone is asking you something and you tell them "this is what the LLM said" you're doing it wrong. If they are asking you they assume you possess some kind of acumen you can use to do better than they can blindly following an LLM and if you can't, well, why are you even involved in this interaction?

Yes, if you are writing a longer piece of communication and you just verbatim ship whatever the LLM spits out, it is important that you note that in the communication. This obviously applies to deepfakes as well.

But this quickly gets slippery. If an LLM generates an initial draft of a paragraph, and I edit it 10 times making mostly stylistic changes, do I need to cite the LLM for that passage? How much personal touch is enough to absolve one of LLM citation? We don't require filmmakers to present a "this is CGI" banner during films (only occasionally in trailers), if a filmmaker uses an LLM to generate 20% of the contents of a film, much in the manner of CGi usage, do they need to add a disclaimer?

The text produced by an LLM really doesn't function the same way text produced by humans does. The cited human is a stable entity that exists, not a stateless machine. You can ask them about their quoted work, look up their history, etc. There is a stable participant in the field with a stable identity that you can talk to. Not so for LLM produced text.

At the end of the day whatever it is your present is coming from you. Own it. If you feel the need to hedge that such and such was from an LLM it probably means that it's bullshit that you'd better rework or at least understand much better before you pass your slop off to some poor coworker

Also, I don't hear a lot of people credit fuzzers. I do hear a lot of people say things like "we are fuzzing the code"'just like they say "we are hammering the nails" and not "wow hammer great job getting that nail in place"

You piloted the LLM, you are responsible for the code. I don't sign my commits with "autocompletions by Intellij". This whole thing is nonsense.

As for the possibilities listed in TFA, I personally think the most common one is people just blaming the tool or hoping to use citation as cover for slop work. It's basically an admission of "I didn't look into this as thoroughly as I myself think I should have to really understand what I'm doing". Don't be that person. If you don't understand what you are doing prompt the LLM with questions until you do understand. It isn't difficult.

At which point i'd argue that you're kind of refuting your own claim as most humans are not brilliant polymaths with early onset alzheimer's.

Even admitting these kinds of comparisons to edge case human mental experience, there are more differences than similarities, and the similarities are misleading and superficial. There are still a lot of differences, even when it comes to the physical structure of the brains neurons as compared to digital neural nets, and i think it's far more beneficial to try to understand these machine in their uniqueness and for what they are than to draw hasty and shallow comparisons. You could be forgiven for that, though, because computing is rife with people who love to draw hasty unjustified analogies for some reason (example: people were already likening the brain to a computer when we didn't even have working implementations of neural nets yet and a computer was literally just a small number of logic gates lol)