Good. For my last GPU is was forced to go with Nvidia since I needed CUDA. ROCm was useless for consumer GPU’s back then. Support has gotten better since from what I’ve seen, so hopefully they continue and my next GPU will be team red again.
Not sure if I fully agree. I’ve seen many people who grew up in luxury and safety who care a lot about the less fortunate in society. There’s also been many instances of 2 siblings growing up in a very similar environment, yet one is very empathic and altruistic and the other one turns out to be a selfish greedy person.
Not that I think you’re completely wrong though. What life you’re born into has a huge effect on who you become. Yet people seem to deal differently with the cards they’ve been handed.
If I look at me personally, I feel like it’s partially genetic, partially in what kind of household you grow up, and also partially your experiences while growing up. I was, for instance, quite “soft” for a guy which meant I got bullied by the more dominant and selfish people. I just wanted everyone to get along and didn’t want to hurt anyone, and because of that I constantly got kicked down. For me it’s turned me into someone who still wants everyone to get along, for everyone to get a chance. Yet it could’ve easily have fallen the other way, where I would’ve let the hate and resentment for other people get to me and turn me into a more self-centered and distrustful person. I feel like this often happens when people fall into the incel or alt-right movements. I think it’s mostly just a combination of how my brain works (genetic) and the support I had at home that kept me away from that. Had I gown up in an environment where my parents would’ve been more “rough” themselves, I’m not sure it’d have gone in the same direction.
After more than a hundred hours (and probably also 50+ crashes) me and my friends managed to finish our co-op Satisfactory session and completed the (for now) last stage of the space elevator. At the end of the game everything was slowly falling apart though. No means of transport was reliable anymore due to the lag. Trains were acting weirdly, hypertubes cannons were very dangerous, standing on drones was a quick trip to the void, and even normal hypertubes had random glitches where you’d randomly fall out of them and get completely desynced. But the factories were impressive, and it was a cool experience.
Now we’re kinda filling the void with some older games. Playing some Splitgate again, which was quite fun despite the low player counts. And also playing some Mario Kart etc via Dolphin. Yesterday we also played Barony for the first time. I didn’t have a full idea what was going on, but one friend had some experience already. We managed to get like 6 floors deep on our second run before saving the game and calling it a night, which felt quite good.
In single player I’m picking up Cyberpunk 2077 again in preparation for Phantom Liberty. My build seems completely messed up tho, and the SPT Grad sniper that I relied on seems to be nerfed and can no longer shoot through walls it seems. Nevertheless the changes seem like an overall improvement to the game, I had a lot of fun with the police already.
Very rare based moment for the UK government. Shows you that even a broken clock is right twice a day.
During my bachelor and master I spent a lot of time with machine learning. Back then things were already moving fast. When I finished my bachelor, I still did my thesis with Tensorflow 1.x, everything was clunky, and transformers were still in their infancy. During my master’s degree I saw it all unfold. I spent most of my time with image generation using GANs and with reinforcement learning. Advances were going rapidly, and development got easier and easier with libraries like PyTorch going mainstream and hardware and software gaining more and more performance. While I was diddling around with generating measly 64x64 pixel images with questionable quality, the field I was working in was improving a lot. GANs became more and more stable, and papers like ProGAN, BigGAN and StyleGAN were pushing way beyond what I assumed was possible.
But surely that’s where it’d end right? Generating high-quality faces, almost indistinguishable from real ones. Even being able to encode them into the GAN latent space to modify faces of real people. And LLMs that could generate real enough sentences to seem somewhat intelligent. I figured that maybe in something like 10 years they’d manage to push this to a level where I couldn’t discern it from a human anymore.
But I was wrong. When DALL-E released I couldn’t believe what I was seeing. The images were obviously fake, but the fact that it could generate images from whatever text you gave it? That was way better than what I assumed was possible. Then came DALL-E 2, Midjourney, Stable diffusion, etc. Way beyond anything I expected, way sooner than I had ever imagined it to be possible. From a consumer GPU anyone could generate images of anything.
And for the first time I felt fear. Maybe this is going too fast. I didn’t have much time to recover though, as next came ChatGPT. I was surprised to see something with such impact being publicly available, and I’m way more surprised that they kept it available. And the advances haven’t stopped coming since. DALL-E 3 just released, GPT 4 exists, and I have no doubt that this will keep going for a while.
Honestly, I’m getting a bit scared of it now. The pace is too high and the effects are too real. These are no longer just fun science experiments, showing some kind of potential. These are now powerful tools to generate fake data that looks real at a staggering rate. In a world were the thruth is already under pressure and many people just believe whatever the social media feed them, we’ve now created weapons of mass disinformation. I feel like we’ve reached a point where most people would not instantly recognize generated images when used “correctly”, not would they be able to identify all generated text. These companies like OpenAI, Google, Microsoft, etc are going (imo) being what is ethical just to win the new AI race.
I hope that regulators will step in to control this, though I’m unsure how that would happen. The research should still be done, the genie is out of the bottle. If we ban LLM research and other countries do not, we’ll still get swarmed with generated fake news. We need better detectors, but I’m not sure that those are possible either. And even if they’d work like 90% of the time, good luck getting anyone to trust them. People will believe generated need stories as along as it fits their opinions.
So honestly, I think that we’ve gone too far and there’s no way to close Pandora’s box. Though I don’t think it’ll end with AI taking over the world, I think it’ll end in a situation where the truth is meaningless and everything can be faked
Same here. I’m cheating though because I’m Dutch, so cycling is really accessible here. I don’t even have a car, so I pretty much have to go by bike.
Running works for me because it’s so easy to start with. For most other sports I have to go somewhere to do it or prepare all kinds of gear. For running I just change clothes, put on my running shoes, and I’m ready to go. The low threshold before I can start, and the low complexity thereafter is what made running stick and what doesn’t make most other things stick.
Ah I see. Weird how they manage to put a fascist message in an otherwise funny meme about chicken nuggies. It really is a dogwhistle then. Unless you know the material it references, it’s just a funny meme.
Maybe I’m missing something, but what part of this is fascist propaganda? It looks like a 4chan meme, but of there’s any fascist dogwhistles in there they’re going right over my head
Something must’ve gone wrong because I’ve yet to receive the mandatory maid outfit
Oh I wish to stop looking at shit memes, but if more than half of them are tankie crap it’s obviously going to degrade the experience
Does your instance federate with Hexbear and Lemmygrad? Because I constantly have those people in my feed and it’s absolutely exhausting. I see more memes about burning the rich and starting a revolution than normal memes in my feed. I’m left-wing, but I’m not looking at memes to constantly be confronted with politics. Unfortunately instance blocking isn’t working well for me yet
I was always taught to eat veggies, so I kinda feel bad every time I don’t. Therefore veggies haven’t ever really been an issue. Fruit is a different issue though. I used to always eat cookies or something a few hours after dinner when I got hungry again. After I decided that they probably wasn’t the healthiest of habits, I replaced that with eating fruit. I don’t even have snacks and cookies in my house anymore most of the time, so I’m kinda forced to alleviate any hungry feeling with fruit.
The joke is that someone who is non-binary doesn’t identify as male or female, yet Spanish is a gendered language and thus ChatGPT provides male and female forks of the word “non-binary”.
As someone who very rarely watches series, this really hits home. I just rarely have the patience to do so, and because of that I also don’t have any streaming subscriptions. I’d rather play games, because the interaction makes them a lot more rewarding to me. To each their own though, other people don’t play games so they’ll have the same problem with me
I’ve hypothesized before that he just bought Twitter to destroy a platform for “woke” people, but it’s still weird reading it out like this. This whole thing reads like destroying one of the biggest social media platforms while wasting billions of dollars is about the same level of investment for him as buying a new phone is for me. Not insignificant, but also something he just does on a whim. And honestly I think his daughter will not exactly like him more after this lol. Rich cunts like Elon buying whatever they fancy just to gain more power is exactly what she’d hate if the article is to be believed.
I can’t really imagine working on any code base that has to actually be maintained and doesn’t have tests. The amount of times that tests have safed my ass at my job are uncountable
Nope I love them. They’re the only moment in the week where I really have a sufficient level of energy to properly enjoy some of my hobbies. On weekdays I’m always a bit too drained to really properly enjoy the more mentally or physically intensive hobbies
Honestly, this gives me strength. 2019 was a year where I wasn’t doing too great. Back then everything seemed bleak, and fun things felt like they were just a faint light in a long dark tunnel. But looking back at that year, I now have plenty of fond memories. So if you’re feeling down now, then think back to the past times where you also felt down but where you now feel nostalgia towards. One day you might look back to your current “good days” in the same light.
A VPN doesn’t really protect you from Google though. They get their data through trackers, which doesn’t get blocked by a VPN. Obviously I’d still not use a Google VPN, because who knows what they’d do with that data
I still want to look into some reviews, but I’ll probably end up buying Phantom Liberty. The 2.0 update has been a lot of fun for me. The rebalancing of weapons, perks, and cyberware has made my build a lot less boring. It was too OP ik the end game, being able to basically kill all enemies without them seeing me. But now my sniper cannot shoot through walls anymore and I constantly run out of quickhack RAM which forces me to use health for quickhacks instead using the overdrive ability.
The police chases are also good fun. It makes the city feel a lot more dynamic. The police will basically engage anything that hurts them. So you can get them to resolve their own NCPD sidequest thingies by leading them into a crime scene with other enemies. They’ll also shoot at civilians or each other. This lead to a peak America moment where a civilian car got launched in a car crash, and because it hurt one officer all the police in the area started shooting this poor civilian car that was flying through the air. It’s been a lot of fun to just drive around and mess with the police system.