ranzispa

I'm afraid we have nothing more to talk about. While I have no idea whether people working in the US military are incompetent or not, I would tend to think they are not. It appears you think the opposite. One of the two of us may be closer to the truth, however I doubt any of us actually knows. It's been a good chat!

I can assure I work much faster. Maybe it is a bit different since I work in research and that's indeed different from working on a large established codebase. Most of my projects are greenfield.

However, recently using Claude code I started many different projects I'd never have approached since I knew it would have taken me months to complete correctly. We are talking about porting file format readers and writers to new languages, then implementing novel algorithms to process such data and optimize it to work on different GPU architectures. Getting a working software takes about a week of work. A publishable cleaned up codebase 2/3 weeks. No way I could do this myself alone in such span of time.

On the other hand, I have students working with me, using AI is definitely not helping them to learn how to code and how to reason. In several occasions I had them showing me a novel huge equation which apparently worked, but then looking at it properly was just over fitting data and they had no way to explain why such an equation should be used.

Fair, but self hosting stuff has that part of self. It is difficult to make it easy for everyone since everyone has a different setup, as such it is mostly directed towards people who are expert in doing this kinds of things or who will dedicate the time to learn how to do it.

The good thing is after you spent a couple days trying to figure out how to make it work, it will work in the future and you already know how to setup more stuff.

Isn't this the requirement of every website with a login system since the 90s? Never used Netflix, was it possible to subscribe without providing an email address?

Also, temporary email addresses exist, as well as setting up a throw away email address; don't see much of a problem with this.

I can see you don't regularly have to upload terabytes of data every day. A faster connection would literally change my work life.

It is common to apply new technologies to fields they have not been used before if the technology seems promising. In most cases if you have to rely on the technology you test whether it works better than your previous approach. I am sure that the US army has resources to do such tests. Indeed, testing in house is not the same as in a real world, but assuming no tests were performed is naive and treating the largest as army in the world as an organization working like a high school student preparing the exam on the night before.

Also, the fact they got a worse outcome than relying on experts

I'm not sure why you say this, I'm afraid this kind of information only the US military has access to. They may diffuse the information because they got great results or as false flag; difficult to know what is real in these cases.

On a broader note, the results of a war tend to follow the better leaders.

War tends to be won by good strategy, tactics do play a role, but without good strategy good technologies are worthlessness. The US lost in a despicable way several wars, they did not do well in Afghanistan, and they were fighting an occupied country with no resources.The current war is a terrible thing happening, but it is not comparable. This is occupying a broken country versus attacking a developed country who has been preparing for war for many years and with a large population, with a very difensible geography. I do not know exactly why the US decided to start this war, this is not public information; however the two wars are not comparable. Even taking out of the equation the alliance with Israel and being brought into other conflicts, attacking Iran is not comparable to attacking Afghanistan. I would not place the blame on the incompetence of the army; indeed generals had plans to attack Iran, as they do for many countries of relevance for their national. The decision to attack is of the government. The job to execute a ch attack is of the army. The idea is that the army talks to the government before the attack happens, if it difficult or impossible the army would stop the government before it happens. I can not know what happened in this case. I don't know whether the army was overconfident in their ability to attack or if the government ignored their advice not to attack. Their war is not doing well, but this is not necessarily the fault of the generals; if they have to fight a war they can not win it is difficult to blame them for not winning it.

We have a different opinion here. I don't think general public has an understanding of the difference between the two things. It may be useful to teach people about it, but I'm not sure how much it would help. People understand generative AI when they hear AI because it's been the only thing they've been in direct contact with. Despite them using ML every day in different aspects of their life this is not something they get to know about. LLMs are different since they directly use the model.

I feel it is good people are getting to know ML and the fact that it is a useful technology; the fact that currently it is associated with LLMs is a side effect. It probably would be better if it wasn't that way, but I do not see it as a big problem.

On the fact that it is polarizing, sure that is currently the case; probably it won't be so in 10 years. That is nothing compared to the past 70 years in which ML has been used and applied in several fields with great success without anyone knowing it even existed.

I work in the field of machine learning as a researcher. There is no distinction. Machine learning is a set of techniques, Artificial Intelligence is a synonym which became popular in recent years.

Generative AI only defines how a model is used, not what the model is. You might as well use a kNN model to do generative AI. Generative AI is machine learning. More specifically it is an application of supervised machine learning.

Some people have started distinguishing machine learning to identify classical models and AI to definite stuff that uses deep learning.

If you want to be specific you may use the name of the specific model, but that's a bit too much for an article title. Otherwise, the term AI is quite appropriate in this context and far more recognized than ML.

And me myself never say AI, I don't like too much that it ingrains some feeling of intelligence; but for these articles i feel it is appropriate.

Sorry, I'm not from the US. Does that include generals and army technicians? In most countries the government does not have direct power over the internal organization of the army, I don't know how it works in the US.

The results of the war do not depend solely on the quality of a single tool or weapon.

The US has better weapons overall and is not winning, thus it is not a way to discriminate whether this technology is useful or not.

I doubt the government directly manages military decisions. I'd imagine the people taking these decisions are in that position regardless of which party is at the government.

I'd imagine they measured metrics of other automated systems and compared with LLM and found out LLM works best.

It's not like traditional machine learning models are suddenly perfect and work in all cases.