There is a guaranteed cap on how far LLM based AI models can go. Models improve by being trained on better data. LLMs being used to generate millions of lines of sloppy code will substantially dilute the pool of good training data. Developers moving over to AI based development will cease to grow and learn - producing less novel code.
The massive increase in slop code and loss of innovation in code will establish an unavoidable limit on LLMs.
Maybe we'll train the llms in our ways of using them, and the next generation of coding assistants will be another layer inbetween us and the code. You talk to the chief engineer llm who in turn talks to its cadre of claude code instances running in virtual tmux. \hj?
But they're not just training off code and its use, but off a corpus general human knowledge in written form.
I mean, in general not only do they have all of the crappy PHP code in existence in their corpus but they also have Principia Mathematica, or probably The Art of Computer Programming. And it has become increasingly clear to me that the models have bridged the gap between "autocomplete based on code I've seen" to some sort of distillation of first order logic based on them just reading a lot of language... and some fuzzy attempt at reasoning that came out of it.
Plus the agentic tools driving them are increasingly ruthless at wringing out good results.
That said -- I think there is a natural cap on what they can get at as pure coding machines. They're pretty much there IMHO. The results are usually -- I get what I asked for, almost 100%, and it tends to "just do the right thing."
I think the next step is actually to actually make it scale and make it profitable but also...
fix the tools -- they're not what I want as an engineer. They try to take over, and they don't put me in control, and they create a very difficult review and maintenance problem. Not because they make bad code but because they make code that nobody feels responsible for.
That is a naive assumption. Or rather multiple naive assumptions: Developers mostly don’t move over to AI development, but integrate it into their workflow. Many of them will stay intellectually curious and thus focus their attention elsewhere; I’m not convinced they will just suddenly all stagnate.
Also, training data isn’t just crawled text from the internet anymore, but also sourced from interactions of millions of developers with coding agents, manually provided sample sessions, deliberately generated code, and more—there is a massive amount of money and research involved here, so that’s another bet I wouldn’t be willing to make.
A cross between a text MUD and an early 2k browser based RPG. Hoping to incorporate many advanced MMO and LitRPG based features plus complex economic, npc, guild, quest, and crafting mechanics. It’s more of a passion project/hobby with no expectation of adoption. It has been very fun to build.
Cool! I really love this sort of project. I've abandoned many similar things :) What do you like most about it? What things do you find fun to implement?
> I, _____, do solemnly swear (or affirm) that I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same; and that I will obey the orders of the President of the United States and the orders of the officers appointed over me, according to regulations and the Uniform Code of Military Justice. So help me God. (Title 10, US Code; Act of 5 May 1960 replacing the wording first adopted in 1789, with amendment effective 5 October 1962).
> I ___, do solemnly swear (or affirm) that I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same; that I take this obligation freely, without any mental reservation or purpose of evasion; and that I will well and faithfully discharge the duties of the office on which I am about to enter. So help me God. (Title 5 U.S. Code 3331, an individual, except the President, elected or appointed to an office of honor or profit in the civil service or uniformed services)
I'll grant you that these militias are well-armed and have sketchy loyalties, but if the US military decided they were honor-bound to defend the Constitution, against Trump, they'd crush the Oath Keepers and their ilk.
Many of the members of those militias are themselves military or ex-military. Which already tells you volumes about how likely it is that all of the US military would "decide they were honor-bound to defend the Constitution". Besides, those guys don't think of what they are doing as contradicting the constitution - quite the opposite! They are very obviously wrong, but that has never stopped fanatics from believing in their creed.
The real question, anyway, is not whether the military will obey Trump's illegal orders, should he issue them. It's whether the military would do something to stop the militias if Trump lets them off the leash with an explicit mandate.
And my concern is that most of the lower-ranking officers and people below them will prefer to sit it out. Because if they do something, and it's not enough, they are all looking at actual treason charges and likely death sentences.
Do you really believe that these lower-ranking soldiers are going to sit idly by while they watch their family members get raped and executed (because that's what lawless paramilitaries do) on the nightly news?
But it won't be their family members, in most cases. It will be some despicable "cultural Marxists" or even more abstract "terrorists". And the rapes? Fake news.
Which does match what GP said. Enlisted swear to defend the constitution and obey the president. No limitation to lawful orders or stated precedence between the too. Commissioned officers don't swear to obey the president
What do you mean "no limitation to lawful orders"? That's what "according to regulations and the Uniform Code of Military Justice" means. Article 92 is pretty explicit about what constitutes a lawful order.
A soldier comes to their own assessment then a court martial decides later on if they agree. Just like every other form of action: do what you believe is within bounds, then find out later if the courts agree
The individual. If your boss orders you to do something illegal, it’s up to you to be sceptical and do your research. That’s the only way it can work without converting a part of the population to drones.
It took me close to 20 years to finally shed the guilt of unfinished software projects.
Side projects can just be about having "fun". Just as going for a walk, or reading a book, listening to music, etc can just be done for fun. I don't have to finish any of these things. I don't have to finish side projects. Fun things don't have to be useful, purposeful, practical, or anything but they should be fun.
I believe Kagi uses most major search indexes as well as its own.
Additionally, I don’t think it’s fair to say it’s more interested in LLMs than focusing on search. I think it’s fair to say they’re interested in ensuring they’re offering a better, non ad-based search replacement.
Disclaimer: Not affiliated with Kagi in any way, just a long time happy user.
The massive increase in slop code and loss of innovation in code will establish an unavoidable limit on LLMs.
reply