In my unqualified opinion, LLMs would do better at niche languages or even specific versions of mainstream languages, as well as niche frameworks, if they were better at consultig the documentation for the language or framework, for example, the user could give the LLM a link to the docs or an offline copy, and the LLM would prioritise the docs over the pretrained code. Currently this is not feasible because 1. limited context is shared with the actual code, 2. RAG is one-way injection i to the LLM, the LLM usually wouldn't "ask for a specific docs page" even if they probably should.
100% agreed on both points. Point 1 relates to https://news.ycombinator.com/item?id=43486526 as well. It's one of the biggest challenges, though maybe it'll automatically get better through models with bigger context windows (we can't assume that though)?
At some point our understanding of fundamental reality will be limited not by how much the physicists have uncovered but by how many years of university it would take to explain it. In the end each of us only has one lifetime.
Comments that explain the intent, rather than implementation, are the more useful kind. And when intent doesn't match the actual code, that's a good hint - it might be why the code doesn't work.
What does "speculative" mean in this case? I understand it as CPU-level speculative execution a.k.a. branch mis-prediction, but that shouldn't have any real-world effects (or else we'd have segfaults all the time due to executing code that didn't really happen)
Cold welding is unintentional, spontaneous joining of two metal parts in vacuum. You don't want that to happen, especially if the parts are meant to move.
Normal welding is intentional application of heat to partially melt two parts at the seam, so that they "mix" in semi-liquid state and become one part when they solidify. Welding may or may not use a third material (solder) to aid the process.
>Cold welding is unintentional, spontaneous joining of two metal parts in vacuum.
Is the only distinction the intention, though?
Because I saw some examples of industrial applications of cold welding, so I'm still not quite getting why cold welding isn't considered welding (I have been googling since my original comment, but not finding anyone making this same distinction).
my understanding, welding is where the metals mix and bond directly. Solder is where they flow and fill the gaps completely, but they're not mixing. According to the webs welding is where both metals melt and mix. Soldering is where only the solder melts. I think you can imagine that if you flow a solder completely between two piece of metal then there'd be no air and it'd basically a really good cold weld. Imagine sticking two pieces of wood together with peanut butter. The PB doesn't actually chemically bond to the wood nor does it merge with the wood.
Cold welding in my understanding is where the metals are very very flat against each other and held togehter by that. like when you have 2 sheets of paper against each other and try to pull them apart without sliding them off of each other or grabbing corners.
In a good weld you actually make 2 (or 3) pieces of metal into one.
No, you don't need to melt the metal to mix; it's just almost always easier to do it that way.
As soon as you cause conditions under which grain boundaries wander, you get welding if the two pieces happen to touch at an atomic/molecular level.
Note explosion welding: you use a shockwave to hold the pieces together while aggressively dislocating grain boundaries. The result is a (very good) weld. Many glues involve welding behavior, especially if they are used without waiting minutes to hours for the bond to harden before loading it.
For example, "contact cement" (polychloroprene glue) works by precipitating a polychloroprene layer from a solvent into the surface pores of both to-be-bonded parts, letting all the solvent dry, and then forcing such prepared surfaces together to cause intimate interaction of the polymer chains on the surfaces to weld into a single layer of polychloroprene that's solvent-soaked into both it's sides (which wouldn't be possible unless the materials are extremely porous).
However, solids are very bad at wetting surfaces, so you will have a hard time getting the needed atomic contact.
Welding isn't really applicable to composites like wood or paper.
> Friction stir welding (FSW) is a solid-state joining process that uses a non-consumable tool to join two facing workpieces without melting the workpiece material.[1][2] Heat is generated by friction between the rotating tool and the workpiece material, which leads to a softened region near the FSW tool. While the tool is traversed along the joint line, it mechanically intermixes the two pieces of metal, and forges the hot and softened metal by the mechanical pressure, which is applied by the tool, much like joining clay, or dough.
Normal welding needs heat to melt the metals. Cold welding happens without heat. Two metal parts will cold-weld on any smooth, touching faces if the air molecules that keep the two separated disappear.
Sure, I mean I read both the wikipedia pages so I've understood that. And I'm not trying to be obtuse here, honestly, but that still doesn't really help me understand why "cold welding, despite the name, _is not welding_".
Your description, and wiki's, both sound like it is welding.
And after spending an embarrassing amount of time last night going down a welding rabbit hole, I have not seen anyone else claim that "cold welding is not welding". I'm pretty sure it is.
Cold welding is basically galling like you get with stainless on stainless under high pressure. Uniformity and consistency is hard to achieve so besides accidentally sticking things that shouldn't be stuck and the lightest duty applications the usefulness is limited.
Humans are not generally intelligent. The adjective "general" in "AGI" does not mean it is equivalent to human intelligence, it means it's above and beyond human intelligence.
I think “general” should be taken to mean “has an average human child’s common sense and causal reasoning,” since common sense and causal reasoning are at some level shared by all vertebrates. It seems like the focus on “above and beyond human intelligence” is how you get AIs which appear to understand algebraic topology, yet utterly fail at counting problems designed for pigeons. It should be scientific malpractice to compare an AI to human intelligence without making any effort to compare it to rat/etc intelligence. (I guess investors wouldn’t lie happy if Sam Altman said “in 20 years I believe we’ll reach ARI.”)
In general tech folks are far too beholden to an instinctual and unscientific idea of intelligence as compared between humans, which mostly uses linguistic ability and surface knowledge as a proxy. This proxy might sometimes be useful in human group decision-making, but it is also how dumb confident people manage to fail upwards, and it works about as well for a computer as it does a rat (though it mismeasures in the opposite direction).
Not at all. Humans are fundamentally limited by our finite statespace and bandwidth. Classifying systems that are able to generalize at least as well as a human but that can exceed those limits as superintelligent is a meaningful distinction.
I agree that "equivalent to human intelligence" is not a robust way to define general intelligence, but humans are a general intelligence.
The "learning" process involves analyzing your codebase's context during PR reviews - we don't train on your data (we even will not save them but only calculate in memory). Instead, we use advanced context retrieval to understand:
- Project structure and architecture
- Coding patterns and conventions
- Dependencies and relationships between components
This allows us to provide more relevant and context-aware reviews while maintaining data privacy (some advanced features still is under developing)
To be fair, once someone has physical access to the machine, them having full access is just a matter of time and effort. So at that point it's security-through-too-much-effort-to-bother.
