Man - that’s wild. Thank you for coming though with a citation - I appreciate it!
Man - that’s wild. Thank you for coming though with a citation - I appreciate it!
a quick web search uses much less power/resources compared to AI inference
Do you have a source for that? Not that I’m doubting you, just curious. I read once that the internet infrastructure required to support a cellphone uses about the same amount of electricity as an average US home.
Thinking about it, I know that LeGoog has yuge data centers to support its search engine. A simple web search is going to hit their massive distributed DB to return answers in subsecond time. Whereas running an LLM (NOT training one, which is admittedly cuckoo bananas energy intensive) would be executed on a single GPU, albeit a hefty one.
So on one hand you’ll have a query hitting multiple (comparatively) lightweight machines to lookup results - and all the networking gear between. One the other, a beefy single-GPU machine.
(All of this is from the perspective of handling a single request, of course. I’m not suggesting that Wikipedia would run this service on only one machine.)
This looks less like the LLM is making a claim so much as using an LLM to generate a search query and then read through the results in order to find anything that might relate to the section being searched.
It leans into the things LLMs are pretty good at (summarizing natural language; constructing queries according to a given pattern; checking through text for content that matches semantically instead of literally) and links directly to a source instead of leaning on the thing that LLMs only pretend to be good at (synthesizing answers).
Thank you for responding! I really liked this bit
with a (decently designed) UI, you merely have to remember the path you took to get to wherever you want to go, what buttons to press, what mouse movements to execute.
I think that’s very insightful. I certainly have developed muscle-memory for many of my most-frequent commands in the CLI or editor of choice.
I agree about Visual Studio as a preference. I’ve used (or at least tried) dozens of IDE setups down the years from vi/emacs to JetBrains/VS to more esoteric things like Code Bubbles. I’ve found my personal happy place but I’d never tell someone else their way of working was wrong.
(Except for emacs devs. (Excepting again evil-mode emacs devs - who are merely confused and are approaching the light.)) ;)
I hope you take this in good humor and at least consider a TUI for your next project.
Absolutely. I see what you did there… 😉
But seriously, thank you for your response!
I think your comment about GUIs being better at displaying the current state and context was very insightful. Most CLI work I do is generally about composing a pipeline and shoving some sort of data through it. As a class of work, that’s a common task, but certainly not the only thing I do with my PC.
Multistage operations like, say, Bluetooth pairing I definitely prefer to use the GUI for. I think it is partially because of the state tracking inherent in the process.
Thanks again!
As someone who genuinely loves the command line - I’d like to know more about your perspective. (Genuinely. I solemnly swear not to try to convince you of my perspective.)
What about GUIs appeals to you over a command line?
I like the CLI because it feels like a conversation with the computer. I explain what I want, combining commands as necessary, and the machine responds.
With GUIs I feel like I’m always relearning tools. Even something as straightforward as ‘find and replace’ has different keyboard shortcuts in most of the text-editing apps I use - and regex support is spotty.
Not to say that I think the terminal is best for all things. I do use an IDE and windowing environments. Just that - when there are CLI tools I tend to prefer them over an equivalent GUI tool.
Anyway, I’m interested to hear your perspective- what about GUIs works better for you? What about the CLI is failing you?
Thank you!
Let’s start a patent troll company that exclusively deals in dark pattern bullshit. Then sue every company that implements any of our terrible patents for as much money as possible. Use the proceeds to bribe lobby congress to pass stronger consumer protection laws.
My wife has hella aphantasia. And facial blindness. She jokes that she would make the world’s worst eyewitness to a crime - she wouldn’t be able to recognize the perp or be able to describe 'em even if she could!
But as it has turned out, she’s been having trouble hearing as she’s grown older. We keep the subtitles on and avoid dining in even slightly noisy restaurants. Anything besides nearby face to face speech is a crapshoot.
So she saw an audiologist who ran the standard battery of tests and said “Your hearing is fine… But have you heard of auditory processing disorder?” Turns out that a portion of our auditory processing overlaps with a bit of our visual processing. For my wife, whose visual systems don’t all work right, her ability to filter noise and extract meaningful sound has started to degrade a bit. Mechanically her hearing is fine. But she can’t process the signal right - possibly related to her aphantasia.
There’s a lot we don’t know yet about aphantasia. Like synesthesia, it’s only been described medically in the very recent past. But if you ever start struggling to understand conversations… Check for auditory processing disorder alongside the rest.
For my wife… well, sge was able to get hearing aids - which look sleek now - and it’s been night and day. Instead of shifting frequencies, these are like noise canceling headphones on steroids. She’s suddenly able to follow conversations in noisy environments that would have been impossible before! So… if any of that sounds (heh) familiar - there’s help available!
Mechanically - both games are puzzle games in the same rough 3d-platform-puzzler vein as Portal. Instead of solving puzzles with teleportation however, you’ve got laser beams and force fields.
On a more metaphysical level, the first game is a philosophical investigation of what it means to be human - to be alive and an individual.
The sequel is a meditation on what makes societies succeed or die.
Both games are fun, the puzzles are just hard enough to be interesting with a sprinkling of well-hidden secrets. But the real reason to play The Talos Principle is if you’ve got an interest in philosophy - the storylines are deeply interested in asking some very big questions. … and they don’t provide answers either - the game poses questions and allows you to answer as you see fit.
If you liked FO3 you’ll like 4.
It’s a lot stronger mechanically than 3 or NV - shooting is a lot less janky and the gun customization adds some great emergent quests.
The Boston of FO4 has its moments - a certain duck pond stands out to me in particular - but aside from Nick Valentine the questlines are largely forgettable.
Still, the core game loop is a lot of fun - go here, blow stuff up, scavenge bits to upgrade your stuff.
As a longtime Fallout fan (came for the isometric apocalypse, stayed for the 3D googie architecture) I still put 80 hours into FO4.
It’s a good fuckin’ game. It’s just competing with the legacy of a lot of other great games in the series.
This is why I loved thee Demon’s Souls approach - single player game with crowd sourced miniboss AI.
If you haven’t played it already - Inscryption - by the same author - is also really good.
… and I should lay off the emdash for the rest of the night.
And it’s currently $5.00 on steam. You know, if you’re price shopping.
Experience.
So… unlike Stable Diffusion or LLMs, the point of this research isn’t actually to generate a direct analog to the input, in this case video games. It’s testing to see if a generative model can encode the concepts of an interactive environment.
Games in general have long been used in AI research because they are models of some aspect of reality. In this case, the researchers want to see if a generative AI can learn to predict the environment just by watching things happen. You know, like real brains do.
E.g. can we train something that learns the rules of reality just by watching video combined with “input signals”. If so, it opens up whole new methods for training robots to interact with the real world.
That’s why this is newsworthy beyond just “AI Buzz” cycle.
Well - I played both and I quite enjoyed Heaven’s Vault as well.
I played HV through twice - once for the story and then a second time to see how far I could alter that story with different choices. My wife even played a third time to try for a really particular set of events.
The translation game in HV goes much harder than Chants’. After the first playthrough, you get longer and more challenging texts to decipher.
Also - there’s no backtracking really required. The game is pretty strict about telling you where you can and cannot go and reacting to what you found or didn’t find. You can cut whole plot lines in HV and it’s no problem.
Which makes it one of the better games for replayablity in my mind.
It is - for sure - slow paced. Almost meditative.
Oh man - I loved WRoEF, but the bathtub segment has ensured I can never play it again.
Ugh - this game!
I loved it. The mechanics of The Scene is still one of the most amazing bits of storytelling I’ve seen in a video game. I think about it frequently when I’m considering how video games can tell stories in ways that movies or books just can’t.
The game as a whole is good, but a little uneven IMO. But I’d put that scene up there with Braid for the sheer impact of storytelling-via-videogame-mechanics.
If you don’t already know… the “corrupt” text in the terminals is where a lot of the semi-secret story clues are - especially in the beginning. If you want to know how to read it, lemme know and I’ll tell you what you need. Otherwise, no spoilers.
That said, the puzzles in the game are pretty consistent throughout, so if solving 3d spatial arrangements of laser beams isn’t fun for you - it’s not gonna get any better.
My most played Steam game is Dishonored, at 127 hours. I have replayed it a lot. A rarity for me, but I really liked that game. Dishonored came out in 2012. It’s taken me 12 years to accumulate that many hours.
Balatro came out two months ago.
I have 93 hours in it.