More training data and token lengths seem to help, given how GPT-4 scores better on a lot of standardized tests than 3 and 3.5 do.
We don’t necessarily need to teach it not to lie, but just to improve accuracy through better training and training data. It (probably) won’t ever be 100% reliable, but what is? Google searches can be inaccurate, same with Wikipedia and other encyclopedias.
Definitely not. Electrek bought a LEV (Light Electric Vehicle) truck like this and it wasn’t. They’re unsafe for roads due to lack of many safety features
That would essentially be at will employment, our current system. “Stopping” at two years is an improvement from our current system because employment (or rather, a salary) would be a guaranteed via a contract. People wouldn’t sign a weekly contract like that if a 2yr contract existed elsewhere
This is a great idea for learning Go! I’ve been trying to pick it up recently and tried to do a CLI for indexing and searching through files. Reading PDFs is such a long hill to climb while trying to learn a new language. A game or something with a TUI sounds like a much more enjoyable goal.
Great article! Enjoyed the argument but I have a few disagreements
I disagree with the response to objection 1. Look for more than a few moments at the right side picture and the shadows of the chess pieces are low quality, the shadows continue off the table, and the checkers on the board look more like backgammon than chess. The AI generated sites that clutter google are easily identified as low quality GPT 3 nonsense.
I’d also disagree that the AI “understands” the physics of the sun, or that Africa has a lot of desert. The prompt includes the desert, so it pulls from pictures of the desert. It doesn’t understand that people have a face when they concentrate either, it finds a picture of someone playing chess then maps an “African fortune teller” onto that person. Oh, it also has to grab photos of an African fortune teller. The AIs used as examples in this article (DALLE, Copilot) are regurgitated humans. What happens when we have new tasks that we can’t train on? They don’t “understand” things, they recreate them.
Maybe I’m just typically bad at estimating the exponential growth. I’m not optimistic I can replace humans with Copilot authored pull requests anytime soon. It might be good at “write a function to convert ISO date time to MM/DD/YYYY” but we’re decades off of “create an API for managing a warehouses inventory.” Either way, being replaced by copilot would require business partners to be able to explain what they want, so I’m not worried.
I’m the same way, not picky as long as I can figure out where to click. I just did the reverse of KDE -> Gnome, I miss KDE here and there but I’ve been able to take the pieces I miss and replace them with gnome extensions.
Gnome was faster to get started with but KDE felt deeper. KDEs UI is denser, which imo makes it feel less “modern” but both KDE and Gnome are usable.
Best thing you can do is to spin up a VM and try it out yourself! Everyone’s different.
I gave an hour presentation on the basics of CSS to my department as a junior. Other asked how I learned it so quickly and where from.
I just read MDN and W3 for like a month while making slides, then played around for a bit with all I learned on an example app. No one had taken the time to do that before I guess?
A fire safe is not a fireproof safe either. Usually they are rated for up to X degrees (my personal one is 1400F) and the papers inside will start to combust around that temp.
> The MetaVerse is a rehash of the Sims (or any other networked RPG game you can pick with avatars) which has now been out for ages, Clubhouse was basically an Internet conference call, WhatsApp is basically a group text/social media rehash, Twitch is basically a video stream and live chat site, and IG stories and TikTok are just a faster way to scroll through mostly edited/reposted YouTube video content, Most web conference tools are all pretty much Facebook with Skype on top of it, And Web 3 often looks like the absolute
How incredibly reductive, evident by the “basically” and “just” minimizations of innovative apps. When twitch was launched in 2011 was it “basically” a streaming platform? Was WhatsApp, with its launch date in 2009 “just” another chat app? Are the content sharing, file sharing, whiteboarding, breakout rooms, panels, background effects, automated transcription of Teams and zoom not a significant improvement over the days of Skype?
Sure, it’s easy to look back 10 years and say X is just Y, cause it’s been around for 10 years!
In college I got the advice that if you couldn’t highlight text on your resume then a bot can’t read it and it might get thrown out. And if that might cause it why risk it?
I think the fact it’s low cost to avoid and high cost if it occurs keeps most people just doing it regardless of whether it’s true or not
Go 50 miles away from your nearest city and look at the sky. There’s a ton of places people live that aren’t light polluted to a level acceptable for amateur astronomy.