Where’s his Bane mask? He’s not prompting as fast as he could be.
I always used to laugh at the way they give instructions to the computer so conversationally, especially building holodeck programs. Like, “how did the computer know to create all these details and characters without being specifically instructed? Who programmed these pre-set commands so that the computer would understand?”
Now, it’s suddenly very realistic.
Yeah, especially when the holodeck goes all haywire and murdery afterwards
Geordi: “Computer, make a holodeck program that would challenge Commander Data.”
Computer: “Make a Sherlock Holmes novel and give Moriarty a comprehension of what exists outside of the holodeck as well as administrative access to the ship’s computer? Okay!”
Oh my god Moriarty broke out of his sandbox and hacked the federation’s computer bank!
That episode of SNW did something similar. Ortega needed a mystery that would challenge even her, so the computer made a fake Spock just to throw her off
Which means the show was 40 years ahead of its time on that one. Incredible.
Holodeck settings and scenarios are open source, and it’s a big galaxy to pull from.
But they aren’t perfect, as the Bynars showed. Without extra processing, a very detailed Minuet turns into just another holodeck character.
(Sad Riker trombone sounds)
Plus it has to do with what is being suggested. When they were trying to recreate the table from the crew being abducted you could tell the holodeck was creating the basics based off their suggestions. If they pull from a book, like Dixon Hill or Sherlock, unless the program was defined in detail beforehand, the holodeck would likely pull from endless historical accounts, photos, etc to fill in the gaps.
And like any learning model, the more a person uses the holodeck the larger their user profile becomes with preferences and such.
Yes, and the ones where they had to make adjustments to what was given made it seem even more realistic. Take the example with Riker. It probably used his past simulations to pull up a typical jazz bar. Then he changes some of the parameters, then a few more. The Bynars monitoring him saw an opportunity and gave him Minuet for his sole audience member.
Still have a huge problem with no locking access. A few minutes later for Picard, who knows what he would have walked into.
The Bynars have all the best LoRAs.
Yeah I also used to laugh a lot when they used those isolinear chips for something they were totally not intended for. Just by rearranging and tapping a few things. As an embedded systems engineer it made no sense to me. In the systems I designed at the time every component had one specific purpose. Sometimes I would get clever and use one component in several different ways to reduce the BOM. Or have a component do double duty, but then it was designed for that (like 4 gates in a single package). The idea you could take out a component, do some magic to it, and suddenly it would do something completely different was just so wrong it was funny.
These days everything is made with whole ass Linux stacks. Like full system on a chip, multiple gigabytes of RAM, tons of storage, often even network access. I recently repaired the door on my dishwasher and the little board that opened the door had a full on ARM core on it. I think it is more powerful than many of the PCs I had in the 90s. I myself have a clock I made using a dual core ESP32 and a simple CO2 particle count sensor for my homeassistant with an ESP8266. With ESP Home flashed it can do so many things using a very small and simple script, with easy OTA updates.
Now the idea of a system built from very powerful subsystems isn’t so far fetched. Just make sure it has access to whatever LAN or CAN they have in the future and it will be able to do it. It might even have some kind of self programming feature, where it configures itself automatically if you kinda sorta get it headed in the right direction.
And I know for a fact they do vibe coding in the future. The holodeck must be vibe coded. Nobody really knows how it works, it breaks all the time. The most important part, the safety protocols are more often offline than online. To quote the great Jack O’Neill: You’d assume a failsafe would be somewhat safe from failing. But no, it fails all the time and somehow the system isn’t designed to simply turn fully off when the safety doesn’t work. You’d think it’s compatible with other systems they have, but somehow it’s not. Smells like vibe coding to me.
I can’t wait for AI models to advance that they can be used for video games and VR games. Like for crafting new equipment visuals or creating new levels or challenges. Generating 3D assets and characters through prompt like you do images should already be possible locally with a fast GPU. And long term it should be possible to be able to generate complete video games at least with specific genres (shooter, RPG etc) and then refine them. Obviously hardware and software isn’t there yet / too expensive. But it could completely revolutionize and change the gaming world.
Makes me remember the games where you “could do anything!”. Spoiler alert they sucked, because there was no “game” in them.
The only thing that will happen, at least in the beginning, is more cash grabs, from AAA to solo devs, good luck wading through thousands of new games a day, all terribly bad.
Ah, so you want to make your own game, for you? Spoiler alert 2, you actually have to know a lot of things to make a compelling game, but it could be an interesting learning experience I guess.
Well you’re not wrong, but I do think we’ll have a lot of people who do want to make good games but can’t today because it’s requires immense capital. Basically AI could become very good at generating “clones” of popular games, which is what most AAA games today do. But then these clones can be further tweaked to become more worthwhile.
That’s where IMO you are wrong, and confound lots of 3D assets with a good game.
There are good games made by 1 person already.
What you are saying could only be true if generative AI can only ever make games worse. Basically if 1 person makes a game and uses generative AI anywhere, the game would always be less enjoyable to the player. Otherwise a “good game made by 1 person” could improve thanks to the generative AI approach I mentioned, and I wouldn’t be wrong.
You can nitpick logic, but if 1 game gets better but is flooded in bullshit games, it’s still a loss for us all.
I can be wrong of course, but I have 10 years of experience building professional games, so I might just have a bit of wisdom to share.
The gist IMO is that if you have the grit to make a game (which 99.99% of people who want to make a game don’t have) then AI isn’t something very noteworthy, and brings its own drawbacks.
If you imagine something like how the crew of the enterprise talk to the holodeck and create games via prompt engineering (e.g. in one episode Geordi and Data tell the computer “create an adventure game in the style of the novels of Sherlock Holmes”) - what would those drawbacks be? And why would that not be noteworthy?
Obviously it would be in a VR environment at best. And rather limited in flexibility at first. And balancing and game design would not become considerably easier. And yes, you’d still need grit and the luxury to at least work part time on it without getting paid. IMHO good games are already drowning in human sweatshop made slop that is controlled by publishers who have the capital.
If people begin pumping even more more games out, the only way really to combat the shit is by increasing the fee to be marketed - It’s 100$ on steam today, but is given back if the game makes 1000$ in sales.
It would be bad for everyone, especially smaller devs from poorer countries if this fee were increased further. But it’s a pretty effective guardrail against bad and miniscule games
That’s just weeding out indies IMO. Any scammer can put up 100€ if he thinks his 1000 shit games will get him more.
I think all this AI and vibe coding is exposing something very instrumental in America…
We all suck at our jobs.
I’m sorry to say…like, I thought it was just me, and it was just imposter syndrome, but nah. Everybody legit sucks.
Not me, I mean. I’m pretty good at my job. I’m sure you are, too. We’re just lazy and doing the bare minimum, and that’s totally fine, I get that. Boss gets a dollar, I get a dime. I hear you.
But like, everybody else. They all suck at their job. I don’t blame them, but it’s more than just not giving a fuck…like they are just legit a bad match for the job entirely. That’s what I’m talking about.
And we all just kinda limped along on mediocrity, and everything just sort of worked.
But now? That misaligned mediocrity has been amplified 1000x. Now we’ve got computer scripts that are trying to do jobs that people 1, didn’t want to do; 2, didn’t know how to do; and 3, didn’t desire to learn how to do better.
And that’s just tearing everything apart.
Programming is especially bad for this because it’s full of people who were told “you should go to school for computers! That’s where the money is!”. So they got into the field solely for the money (not because they had a legit interest in computers/technology/programming), racked up some debt, and found themselves stuck in a career as a code-monkey getting paid a surprisingly less-fulfilling salary than they were lead to expect.
There’s no passion, just overworked, misled, disgruntled people.
Like imagine learning how to write complex programs…by asking google how to do X in Y language and just reading the highest-rated stackoverflow comment.
That’s what’s being churned out.
The one thing I will say boomers did better is CompSci. Practically made the damn thing up as they went along, and was largely made up of passionate, neurospicy nerds. More “Woz”, less “Jobs”. Good for them.
Nowadays, seems so many tech “leaders” make Jobs look like a shy intern.
To rephrase another quote I feel relatable:
“What haunts me is that I am just not good enough in my craft for so many people to be worse.”
We’ve offloaded anything remotely creative to this giant ball of data called AI, and turned most people into robots.
I watched this episode just last night!
Which one is it?
I think it’s “Future Imperfect” from season 4.
Thank you, kind lemmy user!
That’s the one! Complete with bearded Picard.
yes
Something about Romulans?
Kinda, but no. S4E8 Future Imperfect
Sounds like we need an LLM to come up with prompts for us! It’s the perfect solution!
You’re kidding, I know. I’ve done exactly that when I had trouble with planning through an idea.
Yikes
I have ADHD and humans are not always available to help me separate tasks into small packages and report results back to or sort a huge, insurmountable task.
I also have ADHD; I know it can be hard, and I know there are things that make it easier. That doesn’t mean all of those things are good or healthy. Sometimes it’s better to struggle through something on your own than use a tool that is actively harmful for the environment, for the economy, and for our own self-reliance.
If we can’t collectively boycott something as clearly antagonistic to everyone’s best interests as LLMs, then we’ve basically given up on ever fighting back against the downward spiral our world has been tumbling down. Please don’t succumb to the convenience of apathy; we both lived this long without AI, and we can continue to do so.
Sometimes it’s better to struggle through something on your own
Yes, thank you, that strategy finally, after three decades of struggling earned me a short emergency stay at a psychiatric clinic about six years back due to being acutely suicidal and some much needed therapy for three years afterwards. It’s right up there with “just going out and and walking for a bit” and “enjoying the sun”. I’m completely fed up with people telling me to suck it up and I’m not doing that anymore. Nobody in their right mind would suggest not splinting a broken leg because the material used to do so sheds microplastic or something.
Using AI is not the equivalent to splinting a broken leg, it’s the equivalent to abusing opioids because of chronic pain. It helps ease the real suffering we have to endure, but causes even more suffering in return. It’s not worth it, and only serves to benefit the wealthy class shoveling it down our throats.
Hey, everybody has the right to make themselves unhappy or worse as they wish, according to their own priorities, within legal limits. So, you do you.
Dear LLM, what are other people prompting??!
Oy, the spelling made my brain hurt.
It made your brAIn hurt.
They’re putting AI in everything!
Huh, usually Geordi’s the one with the super visor
