egeres @ egeres @lemmy.world

Posts

5

Comments

216

Joined

2 yr. ago

I think you guys are underrating a few of these

• 8 could be useful if you're a spy and want to have your own private documents that noone can read
• 1 you could leave oysters in places suited for eavesdropping, and maybe even coordinate billions of them to tackle down ships etc (?)
• 5 It would be possible to all of the sudden communicate with all humanity in morse code by turning them on and off, you would hit headlines and maybe demand a crazy amount of money in cardano/bitcoin or else you'll just blow up the toast industry. Plus, this morse code thing could lead to a cult where you might have fanatics at your advantage

🥺

I scrolled a bit and there are many like these, yep

I agree with the conclusions of the boomers, but for very different I think long-term AI will produce vastly more harm than good. Just this week we got a headline about google, which is a serious and grown company which already makes billions was up to some fuckery against firefox, facebook has been fined a million times for not respecting privacy and amazon workers have to pee in bottles. To my sadness, all movement against the integration of AI in weapons basically to "kill people" will be very noble but won't do jackshit. Do we think china/rusia are going to give a single fuck about this? Even the US will start selling AI-drones when it becomes normalized. And that's just AI in war, but there's another trillion things where AI will fuck things up, artists will be devalued, misinformation will reach a new all-time high, capchas are long dead making the internet a more polluted place, surveillance will be more toxic, the list goes on

It's reassuring that this opinion is based on many years of experience reading scientific papers, implementing these models and following the trends closely!

See? Contradicting or hiding well-known scientific facts should be unconstitutional

Very noble on their side, but I would bet no government is going to do actual shit about this

Duh, obviously you need to create a windows account to use the clock app 🙄

I just remembered that destin from smarter everyday did a dedicated video about the privacy of this: https://www.youtube.com/watch?v=U3EEmVfbKNs, then, was it complete bull shit?

Is mozilla the only company fighting for privacy?

Up next, windows 13 is cloud-based only, thus requiring constant internet connection

❤❤❤❤ thank you for all the hard work!!!!

Does this count as shrinkflation...?

I partially disagree, the average developer at google is very competent, yet, their work pipelines must be so long and complex that such talent gets somewhat diluted

Also related This new data poisoning tool lets artists fight back against generative AI

No! It's not using an internal exploit, it's rather about finding a way to visually represent almost the same image, but instead using latent features with different artists (e.g, which would confuse a dreambooth+lora training), however, the method they proposed is flawed, I commented more on https://lemmy.world/comment/4770884

Here's the paper: https://arxiv.org/pdf/2302.04222.pdf

I find it very interesting that someone went in this direction to try to find a way to mitigate plagiarism. This is very akin to adversarial attacks in neural networks (you can read more in this short review https://arxiv.org/pdf/2303.06032.pdf)

I saw some comments saying that you could just build an AI that detects poisoned images, but that wouldn't be feasible with a simple NN classifier or feature-based approaches. This technique changes the artist style itself to something the AI would see differently in the latent space, yet, visually perceived as the same image. So if you're changing to a different style the AI has learned, it's fair to assume it will be realistic and coherent. Although maaaaaaaybe you could detect poisoned images with some dark magic tho, get the targeted AI then analyze the latent space to see if the image has been tampered with

On the other hand, I think if you build more robust features and just scale the data this problems might go away with more regularization in the network. Plus, it assumes you have the target of one AI generation tool, there are a dozen of these, and if someone trains with a few more images in a cluster, that's it, you shifted the features and the poisoned images are invalid

Yeah but, it's just one product, the rest of FAANG have multiple gooses that lay down golden eggs

I know that, but why not microsoft?

It's not like I want to endorse them, but to me it seems like they deserved the spot (?) I don't know, back then satya hadn't taken the seat (he was CEO in 2014) and microsoft still was a bit crusty and not so close to open source

I get that at the time, but even back then Netflix didn't have that much valuation compared with the top of the tech companies. Sure, they are very relevant even with the rise in streaming platforms, but as far as I know they only have 1 good product, no hardware or other diversification. I don't really align ideologically with the following companies, but Nvidia, Tesla, Adobe or Microsoft could have taken that spot, they much more valuation according to https://companiesmarketcap.com/, and as I see it, they do more technological innovation

I feel like people just sticked with FAANG because it's kinda catchy, but I think if you take into account world impact/tech development/valuation/size I don't think it makes much sense that Netflix is there