I make heavy use of the "temporary chat" feature on ChatGPT. It's great whenever I need a fresh context or need to iteratively refine a prompt, and I can use the regular chat when I want it to have memory.
Granted, this isn't the best UX because I can't create a fresh context chat without making it temporary. But I'd say it allows enough choice that overall having the memory feature is a big plus.
Grade inflation is common at many schools. And many difficult technical classes grade on a curve, sometimes to the point where you can get an A with an 85%.
But yeah, I still don't see how an 85% average would be a 4.0.
Not that I disagree necessarily, just wondering if there's a consensus that LangChain is too opinionated/bloated/whatever for real industry applications, or if there's some other reason.
Not original commenter here, and not by first hand experience. BUT. I got this kind of feedback from some communities, and I wanted to understand what companies think of this, I asked some dev that works in a company that sells software to enterprise he says that enterprise still use langchain mostly and they are fine with it. On a personal level I agree with the feedback in that langchain has some drawbacks, but at the same time it's a great way to get started.
The recent Apple paper seemed pretty flawed. Anthropic/Open Philanthropy did a rebuttal paper, The Illusion of the Illusion of Thinking.
> Do they reason? No. (Before you complain, please first define reason).
Defining reasoning is the problem. No, they don't reason in the same way as humans. But they seem to be able to go through reasoning steps in some important way. It's like asking if submarines swim.
I completely agree. The difference is, the utility of the submarine and its sales strategy did not depend on convincing human beings that it does swim like them.
The opposite is true of LLMs. Much of the sales pitch of these companies is that these things are capable of some form of reasoning that is in 1:1 correspondence with human reason. Not true. That's what makes this whole segment of the industry snake oil. It's not complete bunk because they do have some utility, but these companies know what they are doing when they use the term reason in all their marketing and papers. They are, in fact, trying to sell you the submarine because it can "swim".
I agree. And if human civilization survives, your concerns about energy and resources will be only short term on the scale of civilization, especially as we make models more efficient.
The human brain uses just 20 watts of power, so it seems to me like it is possible to reach human-level intelligence in principle by using much greater power and less of the evolutionary refinement over billions of years that the brain has.
I agree humans only rarely step outside the circle, but I do have this intuition that some people sometimes do, whereas LLMs never do. This distinction seems important over long time horizons when thinking about LLM vs human work.
But I can't quite articulate why I believe LLMs never step outside the circle, because they are seeded with some random noise via temperature. I could just be wrong.
Yeah, unfortunately I feel like despite all the advances in Unicode tech, my modern terminal (MacOS) still bugs out badly with emojis and certain special characters.
I'm not sure how/when codepoints matter for wcwidth: my terminal handles many characters with more than one codepoint in UTF-8, like é and even Arabic characters, just fine.
`wcwidth` works by assigning all codepoints (strictly, code units of whatever size `wchar_t` is on your system, but thankfully modern Unixen are sane) a width of -1 (error), 0 (combining), 1 (narrow), or 2 (wide).
`wcswidth` could in theory work across multiple codepoints, but its API is braindead and cannot deal with partial errors.
This is all from the perspective of what the application expects to output. What the terminal itself does might be something completely different - decomposed Hangul in particular tends to lead to rendering glitches in curses-based terminal programs.
This is also different from what the (monospace) font expects to be rendered as. At least it has the excuse of not being able to call the system's `wcwidth`.
Note that it is always a mistake to call an implementation of `wcwidth` other than the one provided by the OS, since that introduces additional mismatches, unless you are using a better API that calculates bounds rather than an exact width. I posted an oversimplified sketch (e.g. it doesn't include versioning) of that algorithm a while back ...
Doing that adds a lot of round trips, so you still really need to do the initial estimate.
(also, probing for whether the terminal actually supports various features is nontrivial. At startup you can send the basic "identify the terminal" sequences (there are 2) and check the result with a timeout; subsequently you can make a request then follow it with the basic terminal id to see if you actually get what you requested. But remember you can get arbitrary normal input interspersed.)
Yes, 100% this. And even more so for reasoning models, which have a different kind of RL workflow based on reasoning tokens. I expect to see research labs come out with more ways to use RL with LLMs in the future, especially for coding.
I feel it is quite important to dispel this idea given how widespread it is, even though it does gesture at the truth of how LLMs work in a way that's convenient for laypeople.
I use /foo/bar/ with the trailing slash because it works better with relative URLs for resources like images. I could also use /foo/bar/index.html but I find the former to be cleaner
It bothers me first because it's semantically different.
Second and more importnatly, because it's always such a pain to configure that redirect in nginx or whatever. I eventually figure it out each time, after many hours wasted looking it up all over again and trial/error.
Granted, this isn't the best UX because I can't create a fresh context chat without making it temporary. But I'd say it allows enough choice that overall having the memory feature is a big plus.
reply