I've made that decision before without the help of LLMs so I'm not sure what you're trying to say here. It feels vaguely insulting to our intelligence.
I've made that decision before without LLMs too. If I had been Googling to find possibly relevant material instead of using LLMs to find possibly relevant material, I probably would have bought from Amazon.
With Googling the "figure out what is going wrong" part of solving the problem is more decoupled from the "figure out where to buy this thing" part. The first part involves Googling, looking at a bunch of results, finding a lot are not relevant, trying to refine the search, and repeating probably many times. After that time consuming process when I have finally decided that I needed a new cable I'd probably just go to Amazon without thinking about it.
I always have a little doubt when buying from Amazon because of commingling, but usually not enough to look deeper into it unless the product is something with a high risk of it.
With the LLM instead of Google I upfront described to it a lot of details of my equipment, how I was using it, what symptoms I was seeing, what diagnostic steps I'd taken and the results of those, and why I believe certain things that could cause such problems would not be applicable in my case.
It then finds all the stuff I would have found by Googling, but because it also has way more information from what I told it at the start it can eliminate a whole bunch of the irrelevant results, so I'm starting out way ahead of where I would be after a first Google. A little back and forth and I know what I need to buy.
At that point I'm still at the LLM screen. Since it is right there tossing in a final question about buying from Amazon vs Best Buy is trivial.
I'm not a frequent LLM user. I have yet to pay for any LLM. (I did have a year of free Perplexity Pro that Xfinity gave to its customers a little over a year ago, but when that expired I did not subscribe.
(There's a funny story there--when it expired and they tried to convince me to subscribe, I asked Perplexity if a subscription would be worth it. It told me that considering my usage patterns the free plan was perfectly fine for me and I should stick with that).
A lot of people now are using LLMs instead of or before traditional Google-style searches when they want information. Not just techies or early adopters. The are or are quickly becoming mainstream.
If they are recommending not buying from Amazon that might be something Amazon would want to address.
I might be wrong, but, wouldn't the recommendation to avoid Amazon if you want to be sure come from the massive amount of training data pulled from internet conversations? The kind that would already have been discussing the issue of counterfeit products on Amazon being mixed in with legitimate products from the original manufacturer, since this is a problem that's been going on for, what, at least a decade at this point, right?
The LLM is inherently distrustful of Amazon due to having consumed and trained on a bunch of text that's about how one should be distrustful of Amazon.
Yes, it is common knowledge, but you need to get that information from somewhere in the first place, and why not a LLM?
And sometimes, common knowledge may be wrong, so it doesn't hurt to use LLMs, search engines and other sources to confirm that. Maybe you could discover that Best Buy has a problem with just the product you want, or any other reason. It doesn't hurt to spend a couple of minutes to double check and avoid losing $70.
Beware though, I'm not highly experienced in using and configuring Awesome (especially for beautification), the code is not meant to be shared and it has tons of dirty parts (result of my experimentation), I only built what I needed for my personal use.
Bigger (more public) project built on Fennel is https://github.com/agzam/spacehammer It's for Hammerspoon though, and my guess you're using Linux, not Mac.
One of the aspects I still haven't figured out is the "proper Fennel REPL", it makes certain things less joyful to write or troubleshoot, but even without it, I see some benefits in using Fennel instead of Lua - it's more compact, more FP-oriented, has macros, and it's a Lisp - I can use structural editing.
Disclaimer: I'm the founder of the company that produced this course, but the tutorial is free and the custom assets used in the tutorial are liberally licensed.
A lot of tutorials you'll find will be for Godot 3 so be sure to search for Godot 4 specifically as a filter. Otherwise be willing to take into account that you'll have to figure out how to make the tutorial code work for your version of Godot (which can be an educational experience in and of itself.)
I've done both the official 2D/3D game tutorials in the official documentation, and although the 2D game tutorial wasn't that bad, the 3D tutorial was quite disappointing as a learning experience. It's still not ported to Godot 4, and they use some weird hacky code to achieve some basic gameplay stuff. (And the finished game itself isn't really that interesting...) Even the 2D tutorial leaves something to be desired, since it just dumps heaps of instructions/code at you without explaining you a more general picture first (which wouldn't be a problem for more experienced programmers, but beginners would definitely struggle)
That said... it's still the most comprehensively written Godot tutorial to date, so I recommend at least trying out the 2D tutorial.
gamedev.tv has a godot course updated for 4.0. I haven't gone through that one but their other stuff is really good. Just wait for sale. they usually go down to $10-$15 for a course.
Have you lived in Silicon Valley? The amount that people from other companies talk to each other is incredible. You're using a technology? Oh just hop in the car and talk to the developers about your use case and help guide the development.
You _can_ do that remotely, but having it happen in person is way more reliable
That doesn’t mean that people in NYC don’t talk to each other as much.
Sure it might be easier to hop in the car and talk to developers in the valley. But I don’t just want to talk to developers (and even if I do, not necessarily developers of a tech company). Technology is primarily an enabler for the rest of the world, not just something significant in and of itself.
I can still talk to developers in NYC. But developers are not the only ingredient of a project, company or ecosystem. I can also talk to the best domain experts in finance (different types, not just VC), media, law, health, government, advertising, academia, architecture, etc. all at one place with more ease than in SF. This gives me, a developer, more perspective than I can glean elsewhere.
I agree with the GP that you’ll find a lot of pure tech in the valley. But the applications of that tech all have significant presence in New York in comparison to the valley.
