My thoughts exactly. Growth is a byproduct of quality. Similarly, if the Fediverse grows too much and quality starts to slip, we should also let it shrink until quality comes back. I think our aim should be quality, and anything else is just a side effect.
You’re not wrong. I remember how Bush, McCain, Romney, Obama, Clinton, and others were called Nazis at different points. While it was never really taken seriously then (as it shouldn’t have been), the term has become virtually meaningless. Where the term was reserved for the worst-of-the-worst, for years, it was invoked at the slightest disagreement. Now that there’s a literal Nazi-adjacent person running and getting called out for it, it falls flat.
Building a signup wizard to use that information to select a instance would seem to be the best approach.
That’s actually not a bad idea. I’m not on board with mining contacts, but I think there’s a simple, transparent way to do this that can actually be fun: a personality quiz. Sure, if someone knows what instance to join already, they can override this. But if they don’t, they get like five questions, and then they are matched to an instance.
This is a much better article. OP’s article just shows the author’s surface understanding of how coding works and how well an LLM can actually code. There’s way more that goes into a programming task than just coding.
I see LLMs as having the potential of being almost like a super library. I can prompt GPT, Claude, etc. to write me a custom function that I copy, paste, test, scrutinize, and almost certainly change. It’s a tool that will make someone a more productive programmer. It won’t completely subsume a human’s ability to be creative and put the pieces together.
At the absolute worst over the next decade, I could see programming changing from writing and debugging code to prompting, stitching together, and debugging.
A government contractor overcharging? I’m shocked! Shocked and dismayed! /s
It really is an open secret that all of these contracting companies overcharge the government. Good on her for getting the receipts and doing something about it!
Which is really sad. What “conservatives” are representing isn’t even conservative at all; it’s just regressive, authoritarian non-sense. Conservatives (real ones) have an important part to play in a properly working political system (which we don’t have in the United States). Ideally, we’d have a a progressive party and a conservative party; the former looking for reforms anywhere they can, even to the point of burning down the system, and the latter urging caution and ensuring the stability and safety of the system. Instead, the U.S. has a conservative party (Democrats) and a fascist/regressive party (Republicans). So, instead of moving forward with caution, we’re moving backward.
I started using it to jump-start coding projects by writing the boilerplate that I could then fill in and focus on things that were more important. To put it in perspective, I was able to produce a prototype for a client in about two weeks rather than a month to a month and a half. I’d never use it to produce an entire project out of whole cloth, but I’ve found that it’s super useful for getting the big pieces done quickly so that I can focus on fine tuning and making a product better. I also started using it for a starting point in professional letters and other such mundane writing activities of work. So far, it’s been great for giving me a starting place that I can improve from while making my work faster.
So, I’m not surprised that those using LLMs have improved their productivity.
After leaving my (very conservative) home, and going back every so often and talking to my family, I think this is it. Especially for the older ones, I get a sense of regret from them that they didn’t live how they wanted or couldn’t do what they wanted. Since they don’t want to blame themselves and/or the culture, they blame their vilified minority du jour. It’s sad really. Going home and talking to my parents always just makes me sad about how I can tell they have some pretty big regrets about taking the “expected” path.
Yeah, you’re right on a lot of chatbots just being paraphrased responses from the support database, but for a lot of people, that’s all they want or need. There are a great number of people who just don’t want to read the entire article to find their answer. For that, I don’t really mind chatbots because I get the use case. What I hate is when there isn’t an option to go to the next tier of support without going in circles forever with the stupid bot.
The original paper itself, for those who are interested.
Overall, this is really interesting research and a really good “first step.” I will be interested to see if this can be replicated on other models. One thing that really stood out, though, was that certain details are obfuscated because of Sonnet being proprietary. Hopefully follow-on work is done on one of the open source models to confirm the method.
One of the notable limitations is quantifying activation’s correlation to text meaning, which will make any sort of controls difficult. Sure, you can just massively increase or decrease a weight, and for some things that will be fine, but for real manual fine tuning, that will prove to be a difficulty.
I suspect this method is likely generalizable (maybe with some tweaks?), and I’d really be interested to see how this type of analysis could be done on other neural networks.