No discredit to R.E.M. but my world's been ending for over a decade and I feel like dogshit constantly. Nobody told me the apocalypse would be heralded by the dumbest fucking cryptobros and AI prompters the world's ever seen.
I wonder if the nft tech could be used to mark everything real with a proof of authenticity. Then those things that aren’t on the chain would be considered suspicious and you would have automated green border around any content that is authentic and red for the outside of the blockchain
How could that help at all? Seeing as the blockchain would have no way of telling the difference between human and Ai text, and if you could find a way to automatically verify that in way way that was so efficient you could expect all the text uploaded to the internet you could just run that program locally and not be beholden to people paying a fee to post anything to the internet.
I can’t imagine any sort of verification system not being completely overrun by bots/people on fiver/ mechanical turk immediately unless you tied it to meatspace IDs in an know your customer sort of way, in which case you would definitely need a central organization to do said verification, which eliminates any possible need for a blockchain as said organization can just use a faster, far cheaper, and most importantly for this application editable database.
More to the point, no one doubts that an article published by one organization was secretly published by another, but rather that they secretly used AI in the writing process, which also negates the system because that organization is never going to tell you which articles are done by AI, and any sort of reporting system for the entire organization or a specific author is just going to be immediately and constantly used to review bomb.
Well, whatever the solution to this problem is, I'm fairly sure "put a blockchain on it" isn't going to be it. Distributed ledgers do potentially have some uses, but using them to carry "proof of humanity" information doesn't make much sense
So again deleted my comment accidentally omg. I very like this idea. It’s burdensome but every phone already has simillar scanner. I wonder about privacy though. Can it be anonymised like monero?
I imagine it as an optional feature and then verified people can opt in to only show other verified humans in some kind of next gen web similar to fediverse. Then two layers form naturally one of verified humans and the good old internet Wild West.
Because it's a centralized system owned by a sociopath billionaire gathering unchanging, personal details about swaths of the population using ye olde "for the greater good" adage as the justification. You'd have to be a special kind of fool to go along with it.
The AI will become utter garbage, mark my words. It will complete the downfall of the internet though first. What will come after internet? Hard to say maybe we will exchange culture in VR? Probably at some point every kind of forum will be unreadable from automated marketing and propaganda. Reddit already is. Lemmy is too small… yet but it will crash even harder if it is noticed by the wrong people. 2 years tops to enjoy this medium
Though I admit I am somewhat of a doomer inclined so maybe there is a way to create alternative that cannot be poisoned
Well, for many publishers the choice is either ads or paywalls. The fact that people feel entitled to get everything for free is a part of why things are going to shit, because ads bring with them a whole slew of perverse incentives (eg. optimizing for ad views instead of content quality)
The paywalls restrict the flow of quality information, which happened before LLMs started scraping the web. If you don’t have money to spend on all of these news subscriptions you aren’t allowed to educate yourself. It’s class-based gatekeeping, plain and simple. They could tactfully include ads, but no one ever tactfully includes ads. They introduce pop-ups, fullscreen banners, interjections every 25 words, or the best is the articles that are just slide shows that take you through 30+ webpages.
Edit: I’d also like to point out that this article already has an ad at the beginning. So they are still making ad revenue even if they aren’t giving you complete access.
This is pretty patently hyperbole; I've run into many sites, including news, with non-intrusive ads.
Whether it's class-based gatekeeping is another matter entirely. For-profit media employees have to eat too, and in the current economic system most can't just give people access to content for free without any sort of monetization mechanism and with a voluntary subscription, because that'll very often lead to income dropping off a cliff. Unfortunately people are very loath to pay for online services except for some more niche cases like the Fediverse where instances run on voluntary donations – although I've seen a couple of moderately popular instances struggling with upkeep being higher than what people are willing to donate (and it's not just services either; open source developers face similar issues.) In some countries we at least have public broadcasting companies, although eg. here in Finland the current extremist right-wing government is looking to reduce its funding by quite a bit and possibly even entirely dismantle it if they get their way.
While I definitely agree that news should be available for free, railing against a for-profit publisher's paywall is, frankly, myopic; like it or not, in the current system even content producers have to make a living. None of us really has a choice in whether we want to live in this system or not
Advertising, by design, is intrusive. It’s fighting for space in your mind whether you want it to be there or not. We can shelve that topic because it’s a side item here.
The difference between making a big deal of nothing and being completely on-topic is that the article itself goes into the responsibilities of publishers and platforms, how they have a responsibility to make the internet a better connected, more human-friendly place. You don’t see massive sources of misinformation locking down their content, but you will definitely see potentially credible sources of information doing that. It’s counter to the premise of the article entirely.
I don’t believe it’s myopic at all to point out that it’s backwards to expect the internet to thrive when quality information isn’t readily available. Sure you can use a different search engine, seek out free content and resources, all of which require an in-depth dive to find anything worthwhile.
The topic of this post is why the internet is dying, and while I recognize people need to make money to eat I think these news media sites are more than capable of providing for their employees with or without a paywall. Megacorps like Google, Meta, and Microsoft having control over what gets the most clicks is definitely contributing to rapid enshittification. Especially when they’re sending most traffic to articles that either have a paywall or a steady feed of bullshit.
The question is how do you expect quality information to be produced if it isn't paid for? I think it's terrible we have to think in those terms but as the other person said, that is reality.
Linux is a prime example of quality that isn’t paid for. No one forces you to pay for Linux, you can of course support the maintainers and donate, but it’s not a for-profit endeavor.
Linux is the result of a massive number of people working at their own paces with no deadlines and no expenses other than time and the computer they already own, as well as foundations where people get paid and pay others do tasks. Lots of private companies are also involved, and they exist because of profits.
Quality, relevant journalism has hard costs associated with it and has to move very fast. I'm not even talking about the twitter blitz that leads to sloppiness. I'm saying any and all breaking news. How do you plan on getting any on the ground reporting in Gaza?
What you are suggesting would mean that only those who don't need an income can participate in the endeavor. Which unfortunately is also the case with Linux - big contributors have to stop all the time, projects die regularly, because "life gets in the way." It just shifts the problem.
Open source programming and journalism have some parallels I'm sure but this comparison just doesn't work on many levels.
How do you propose these "open source journalists" make a living? Corporate grants or straight-up corporate jobs just like a huge chunk of Linux development, landing us right back at square one, if not even somewhat behind it? At least independent media exists nowadays, but if the assumption is that all news has to be freely available, like acastcandream said that'd just lead to journalism being very effectively locked out as a career path for anyone who's not independently wealthy or somehow able to make people actually donate or pay for a subscription despite the content being available for free – and that hasn't worked out too well for most publishers so far.
The largest code contributors to Linux are corporate contributions
Regular people who contribute to OSS do so as a passion project, as a hobby, and have other unrelated jobs that pay the bills. Those people still have to make a living, they're just not doing it from their software contributions. Journalism isn't a hobby and you can't work a day job and still be an effective journalist. News orgs don't come together as hobby projects.
I'm not defending advertising. I hate it and think it's ruined the web. I'm just addressing the analogy here wrt Linux.
They could tactfully include ads, but no one ever tactfully includes ads.
Because they don't work outside of basically podcasting. And even then many shows stretch "tactfully."
Additionally, over 60% of American internet users use an adblocker. The Atlantic as a US publication relies heavily on US citizens. They didn't create that situation, but they have to experience the ramifications.
So "tactful" ads are not an option. You don't want obtrusive ads that unfortunately are the only ones that vaguely work. You don't want to pay for it directly. If they went government funded or something it would be used as a cudgel against them forever (look at how NPR gets shit on it randomly, which basically gets a fraction of its funding from government grants and not even formally from the US budget). So functionally no ads, no selling, no subscriptions, no government funding.
I legit have been considering buying a minidisc player, just for the sheer cool factor of them. Sometimes truly special form is lost as function evolves.
minidiscs should have been the standard CDs became. no one would have considered holding a 3.5" floppy disc gingerly by the edges and placing it into a tray to read. the case is critical.
however, the industry realized people were replacing their scratched cd's, so they'd sell 2, 3 of the same disc to the same person. Source: worked at a record store in the hight of the CD age when mini discs were all but essentially snuffed out
I mean, maybe not the mini disk specifically, but yah, a cartridge system for CDs would have been better.
Mini disks are super cool but they’re a lot more materially demanding than a CD, CDs being just aluminum and plastic, where as a minidisc has some truly wacky elements in it’s make up to get the magneto optical and curie point to work.
I mean you clearly know way more about it than me.. but yeah some kind of carrier/cartridge protecting the disc and we'd probably still be using CD-W-RW's
The mini disk was a truly weird system. Half way between a cassette and a CD. CD used a laser to to reflect off bumps(or dyes in some varieties) on the disk to get a signal, and a cassette would use a metal head to detect magnetization along the tape to get a signal.
The mini disk used a laser to read the magnetization around the disk. Essentially the magnetism would change the polarity of the light as it bounced off, and by measuring what the polarity of the reflected light is, the device got the signal.
Writing to the disk was also wild, as unlike the cassette, the magnetic field of the disk couldn’t just be changed by putting it next to a strong magnet like. Instead, it had to be heated up before the magnetism could be changed, this heating was done with the laser, and was very precise compared to a cassette’s method. This meaning way more information could be squeezed on to the disk than on a cassette.
There were several attempts to chain CDs to caddys, kind of like Laser Disc none stuck.
There were also Zip (250MB) and Jazz (1GB) Drives that were pretty amazing for their day. Unfortunately media was expensive. Jazz was pretty great for backups though.
money! just production costs I'd say. I took a 3D design class forever ago and they made us get Zip disks, and if you wanted to work on stuff at home that meant you needed a zip drive. My dad (RIP) bought me a drive and I'm somehow still touched as it was a reasonably big purchase at the time for a slacker 18 year old doing 3D design at a community college
Not a 3D designer now, finally went back to college for something else in my 20s.. anyway, rip my dad and rip the zip drive, they were cool. 100mb wow
My wife really needs to watch the Ravens game this weekend. From California, because not everyone is a fan of their regional team, and apparently if you are a fan of a faraway team, that means you are a super-fan and should be charged extra money.
I, personally, am not a sports fan, so I’m happy to watch things at later times and dates than the very moment when they’re happening. But knowing what’s happening RIGHT NOW is important for those who are True Believers™
Is there a fast way to surf the streemz like a gentleman of fortune, or do I need to pay my dues and do the research on my own?
Many members of the team, including a growing contingent fearful of the existential risk of more-advanced AI models, felt uncomfortable with how quickly GPT-4 had been launched and integrated widely into other products.
GPT-4 and anything similar isn’t going to pose an existential threat to humanity.
Eventually, yeah, there is probably a possibility of existential risk from AI. I don’t know where that line ultimately is, and getting an idea of that might be something important for humanity to figure out, but I am pretty confident that whatever OpenAI is presently doing isn’t it.
Same reason that Musk and his six month moratorium on AI work doesn’t make much sense. We’re not six months away from an existential threat to humanity.
I think that funding efforts to have people in the field working on the Friendly AI problem is a good idea. But that’s another story.
Being an existential threat is a much higher bar – that’s where humanity’s continued existence is at threat.
There are plenty of technologies that you could hypothetically put somewhere where a life might be at stake, but very few that could put humanity’s existence on the line.
It's the same situation, just writ large. Dumb human decisions to put AI where it shouldn't be. Heck, you can put it in charge of the nuclear missles now if you want to. Don't. Though. That'd be really, really stupid.
Part of my knee-jerk dislike of the AI hype is that it's glorified text completion. It doesn't know shit. It only knows the % chance of your saying the next word. AGI is not happening anytime soon and all this is techbro theatre for the sake of money.
Anyone who reads a wall of bland generated text and thinks we're about to talk to god is seriously mistaken.
I’m much more worried about the social implications. Namely, the displacement of workers and introduction of new efficiencies to workflows, continuing to benefit only those who are rich and in power, and driving more of us towards poverty.
It’s not an immediate existential threat, but it’s absolutely a serious issue that we aren’t paying enough attention to.
How did the industrial and information revolutions work out for us? Sure we live lives of convenience, but our entire existences have been manipulated into making the rich richer.
Looking at long and short term trends in the wealth gap, I have absolutely no faith that this will go well.
You do realize that a lot of people are already being displaced by AI right? These are not “unskilled” jobs either. For e.g. the illustrators who used to get jobs probably spent thousands of hours to get to that level
AI is already taking video game illustrators’ jobs in China
Nothing about this is safe. It’s easily the worst misinformation tool in decades. I’ve used it to help me at work, GPT-4 is built into O365 corp plans, but all the jailbroken shit scares the hell out of me.
Between making propaganda and deepfakes this shit is already way out of hand.
Stuff without the guardrails, stuff that’s been designed to produce porn, or totally answer truthfully to queries such as “how do I build a bomb” or “how do I make napalm” which are common tests to see how jailbroken any LLM is. When you feed something the entire internet, or even subsections of the internet, it tends to find both legal and illegal information. Also the ones designed to generate porn have gone beyond that boring shitty AI art style and now people are generating human being deepfakes, and it’s become a common tactic to spam places with artificial CSAM to cause problems with services. It’s been a recent and long-standing issue with Lemmy - people like Exploding Heads or Hexbear will get defederated and then out of retaliation will spam the servers that defederated from them with said artificial CSAM.
I like copilot but that’s because I’m fine with the guardrails and I’m not trying to make it do anything out of its general scope. I also like how it’s covered by an enterprise privacy agreement which was a huge issue with people using ChatGPT and feeding it all kinds of private info.
Almost everything you said, with the exception of AI CSAM and suicide prevention, can hardly be considered a serious issue.
What’s wrong with searching for how to make a bomb? If you have the wish to research it, you can probably make a bomb just by going to a public library and reading enough. The knowledge is out there anyway
The best part of Open AI’s self professed goal to make an AGI is that the more we learn about LLM’s the more it becomes clear that they inherently can never bridge the gap to AGI.
One would almost think the constant complaining about mythical dangers of AGI might be a distraction from the real more mundane dangers LLM’s pose here and now like exasperating bias, making mass misinformation easy, and of course shielding major companies from accountability.
Or the other option is that it’s just marketing, look at how scary our totally real product is, look how fast it improved when we went from a medium sized dataset to the largest that will ever be possible, don’t ask questions like why would a autocomplete that has been feed the entire internet actually help our business, just pay us and bolt it on to whatever you can.
Scripts and automation do what thier programmed to. There are bugs and mistakes, but you can theoretically get something programmed right. LLM’s generate text that looks like a human language. If they were just getting used to make up random bullshit it wouldn’t be a problem, but there are few applications where random bullshit is actually beneficial.
Thouse all sound like things where it might be really bad if it injects untrue information, and with an LLM, by definition it has no understanding of what it’s summarizing. It could be especially bad if the people useing it actually trust what it outputs as facts about what was fed into it, but if they don’t and still check the source than what’s the point.
If I hand someone a set of bullet notes and ask them to send out a notice in writing to the company. They are going to convert those notes into paragraphs and sentences… Not just send out the notes.
Also MS already has a module for teams that will take the conversation transcript, and output action items based on the conversation… It is like having a note taker during the meeting. www.youtube.com/watch?v=N1gpkk-MwpY
Oh, I'm sure they will. That is not, in the slightest, the same as caring about said implications in ways that mean that the species won't get murked, though.
They believed that the AI safety work they had done was insufficient.
Considering that every new model seems to be getting worse for anything but highly sanitized corporate usage, I’m not sure that I want more AI safety …
For my usage, I use Chat GPT 3.5 turbo with the march checkpoint because I can’t get the current one to stop moralizing about bullshit instead of doing what it’s supposed to (I run two twitch bots with it). GPT4 used to be okay there, but the new preview is now starting to have the same issue with more frequent “I can’t do that Dave”-style answers, though it’s still mostly circumventable with enough prompt massaging, but it is getting harder.
In a year, I don’t see anything but self-hosted models usable for anything not corporate glitz if trajectories hold, so fuck all that AI safety.
On top of all of this, those efforts to tame and control outputs from the developer side could be abused to simply appease investors or totalitarian markets. So we might see a Disneyfication like we‘re seeing on other platforms like Youtube with their horrendous filters, spawning ridiculous terms like „unlifed“. And just imagine the level of censorship we‘d see if they ever try to get into the Chinese market because clearly, the ‚non‘ in non-profit is becoming more and more silent.
It’s already easy to self host and we’ve optimized LLMs to run locally on not much serious hardware after we’ve trained them; I have GPT4ALL set up on my machine and it runs everything locally with my processor, no GPU or anything. Some of those datasets are uncensored, and I’ve seen what Stable Diffusion can do for image generation.
I tend to use the GPT-4 built into Edge with my O365 corporate plan, because it suits my needs better for day-to-day challenges. It can still audit code and summarize things, which is all I really need it to do here and there.
I do not believe any 7B model comes even close to 3.5 in quality. I used LLama V1 64B, and it was horrible in comparison. Are you really telling me that this tiny model gives better general answers? Or am I just misunderstanding what you are saying?
I didn’t say better, I said comparable
And faster, without handing over my data and conversations for monetization
Given the locally hosted benefits, and the ability to go to chatgpt for any answer minstrel gives that doesn’t satisfy you, makes it strong competition to chatgpt as the default tool
Hosting it yourself also means you can swap llm’s out based on context and what they’re trained on. Highly tuned models perform better than chatgpt at the things they are meant to excel in.
I’m currently trying to show on the Website Beehaw, that certain LLMs are far superior in writing than others. Examples of what bigger models do better than smaller ones: *
Mistral-7B-Instruct-v0.1
ntire articles* vs HeadlinesDescriptions vs Product titles *Bul
GPT 3.5-Turbo doesn’t support completion as it’s for chat, so I used an even worse one, text-davinci-003 which is far behind state of the art.
Bigger models are able to handle more complex and detailed tasks with ease
Bigger models are better suited for natural language understanding and text processing
Bigger models are able to learn more accurate representations of context, thus improving the precision of the output
Bigger models can process data more quickly and efficiently, saving time and processing power when large volumes of data are used
Bigger models can better recognize more subtle nuances in language, which allows them to produce more accurate results
Bigger models are able to use more sophisticated algorithms, resulting in a more comprehensive and deeper understanding of the data being used
Mistral 7B might be okay for some very specific cases, but it’s not comparable to proper models at all.
edit: gave it a second chance, it’s a bit better (at least no complete nonsense anymore), but still terrible writing and doesn’t make much sense
Paraphrasing The ability of a language model to generate text that has a similar meaning to the original text is called paraphrasing. This is a very common problem in natural language processing, and many LLMs are designed to be able to paraphrase text. However, there are some LLMs that are particularly good at paraphrasing, and these models are often preferred over smaller models because of their ability to generate more varied and unique text. Examples of LLMs that are known for their paraphrasing abilities include GPT-2 and transformers. These models
theatlantic.com
Hot