Welcome to Incremental Social! Learn more about this project here!
Check out lemmyverse to find more communities to join from here!

Adalast

@Adalast@lemmy.world

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Adalast ,

Not all heroes wear capes. Seriously, this man is now enshrined in my mental halls of great men along with John Bannister Goodenough, whom I just discovered died last year. I'm going to go be heartbroken now.

Adalast ,

Just train a couple LoRAs and you can add in the vectors needed for doing whatever you want. "Circumventing" "safeguards" in this case means using a feature built into the system already. I have been making AI porn from the start. The day in installed their local framework for using it and got a Rick Astly image instead of anime tits I went and found their censorship function and neutered it. It is just python code, super easy to do.

Adalast ,

Yo ho ho my friend. Yo ho ho.

Adalast ,

My hot take, in the digital age, all direct marketing should be opt-in with the platform. Opt-in for industries with the ability to ban specific advertisers.

Adalast ,

As an American and avid rights understander, it is not the 5th Amendment which this risks violating (which you did cite correctly), but the 4th Amendment, which guarantees protection from undue searches and seizures of your person, property, or effects. This is the whole reason for the warrant requirement and the reason you hear us bitching whenever something comes up that lets police or agents of the government acquire non-public access to information or property in a warrantless way.

An example: the police are investigating Mary's death and suspect you of having planned the murder in the Notes app on your phone, so they want to get into your phone. Without a court order (warrant), you have to give them permission. With the court order, you must give the passcode and/or unlock the phone.

Now, at this point, if your passcode happened to be 'I killed John02&' you could argue 5th Amendment protection because divulging the information would incriminate yourself in the crime, or a different crime.

Adalast ,

I have neither the emotional nor mental bandwidth to read up on this, too stressed being poor. Does anything say what is happening to the money?

Adalast ,

Can we take a minute and stop to assess where Adobe is obtaining its training data? Everyone is all up in arms about the OpenAI devs scraping DA and such, but Adobe is 100% training on the entirety of Behance and the Adobe Cloud. Things that are not public, our personal files that we never intended others to be seen. Our private albums of our children, or our wives/husbands/partners, or parts of NDA restricted projects that are stored in Adobe Cloud automatically that are supposedly not in violation of our NDAs.

Where are the pitchforks? Where is the outrage? This is 1000x worse than some desperate AI engineer staring at a publicly visible and available training set that is already tagged and described in detail that was begging to be used. People lost their shit over that one. Why does Adobe get a pass?

Adalast ,

That's fun, glad to see they are paying people now. I didn't see in there when in the multi-years long process it takes to develop tool-sets and train checkpoints they paid for the rights to create derivative works. The article is dated a few days ago and it is present tense. They are NOW paying. The AI is trained. The tool is built. It takes tens of thousands of images to train a generative model from scratch, I would expect decades of footage for a video model. So if the model is trained, and them paying is new...?

Also, they don't have to ask, or pay... They already have the rights for all content stored in Creative Cloud (EULA Link).
Adobe Creative Cloud EULA

Legally, an AI training is a "derivative work", so I would need a letter from the lead engineers on the AI dev team at Adobe, signed by every dev who has worked on it, stating that they only used paid training material at every stage of development of the tools, disseminated separately from any official Adobe channel before I would believe that the greedy gaping maw that is Adobe did not just use the millions of images and thousands of years of footage they have legal right to use that THEY are actually PAID for. They know they can pay now because it is a drop in the bucket compared to the Creative Cloud fees and is great PR and an even better smokescreen. There is precisely 0 chance they are going to receive enough good, usable footage through this program to train an AI from scratch.

Adalast ,

Their Creative Cloud license

That little "derivative works" bit in the middle gives them license to use the files stored in Creative Cloud to train AIs. So yes, they are using their data sets that they have license for. It just happens to be our data that they took the license on and we paid them to do it.

Adalast ,

As much as I am anti-censorship and hate all of this, there is no "Freedom of Speech" on social media platforms. They are private companies and are allowed to use any restrictions that do not fall into violations of the very few laws which restrict how companies can treat customers. In the USA, "Freedom of Speech" only guarantees that government agents and laws will not restrict it, and even that is not absolute.

Now the laws and policies about it need stripped along with all non-symmetric indecent exposure laws.

Adalast ,

Already happens. Facebook and other message systems have content filters built in to their chat systems.

Adalast ,

Way to strawman my man. SMS is not the only mode of communication and SMS and cell phone communication in general fall under a whole litiny of laws because they are considered a utility along with landlines. This extends the constitutional protections for the first and fourth ammendment protections to them. Your initial suggestion was a fallacious argument to start because utilities are not wholly private corporations, and thus do not qualify for the initial discussion. I tried to save it by suggesting that alternative means of communication which are utilized that serve the exact same purpose as SMS and telephone calls but are controlled wholly by private corporations DO fall victim to arbitrary censorship and it is allowed because they are not subject in their business dealings with consumers by any state or federal oversight.

Adalast ,

Yup, as I mentioned in my other response to Mr. Strawman that is the difference between a private corporation and a utility. Utilities ARE subject to the first and fourth amendment protections because they are a strange hybrid between public and private.

Adalast ,

Eh, its more of an oligopolistic cohort, but yeah. There is a strong argument to be made to classify all communication technology broadly as a utility under the telecom umbrella. That way it can cover all past and future technologies.

Adalast ,

The example of stubbing ones toe is a strawman. It is levying a rhetorical argument which has nothing to do with the topic at hand. It is also hyperbolic and sartorial.

Taken from excelsior.edu on the topic of sreawman arguments.
https://lemmy.world/pictrs/image/1b5d8240-94fb-4f0a-8e9f-5c6551921ef3.jpeg
Distort: equate it with a non-topical example.

Exaggerate: use hyperbole in the example to describe a link between stubbing one's toe and deaths in car accidents.

Lack of engagement with my argument: there was no support for your assertion that the companies I mentioned do not engage in phone/textual communication when all of them work on the phone, are able to make voice calls, transmit text between phones, and are able to be set as the default SMS apps for a phone, which then subjects SMS communication to thier monitoring and filtering.

I know it subjects them to filtering as I once had Facebook Messenger set as my default SMS app and attempted to SMS a friend a porbhub link and FBM said that the message failed to send. Non-pornhub links worked just fine, but they filtered my porn message, and it was on SMS.

I know precisely what a strawman argument is. I made a good faith response, you did not.

Adalast ,

The example of stubbing ones toe is a strawman. It is levying a rhetorical argument which has nothing to do with the topic at hand. It is also hyperbolic and satirical.

Taken from excelsior.edu on the topic of sreawman arguments.
https://lemmy.world/pictrs/image/1b5d8240-94fb-4f0a-8e9f-5c6551921ef3.jpeg
Distort: equate it with a non-topical example.

Exaggerate: use hyperbole in the example to describe a link between stubbing one's toe and deaths in car accidents.

Lack of engagement with my argument: there was no support for your assertion that the companies I mentioned do not engage in phone/textual communication when all of them work on the phone, are able to make voice calls, transmit text between phones, and are able to be set as the default SMS apps for a phone, which then subjects SMS communication to thier monitoring and filtering.

I know it subjects them to filtering as I once had Facebook Messenger set as my default SMS app and attempted to SMS a friend a porbhub link and FBM said that the message failed to send. Non-pornhub links worked just fine, but they filtered my porn message, and it was on SMS.

I know precisely what a strawman argument is. I made a good faith response, you did not.

Fake Photos, Real Harm: AOC and the Fight Against AI Porn (www.rollingstone.com)

In 2023, more deepfake abuse videos were shared than in every other year in history combined, according to an analysis by independent researcher Genevieve Oh. What used to take skillful, tech-savvy experts hours to Photoshop can now be whipped up at a moment’s notice with the help of an app. Some deepfake websites even offer...

Adalast ,

Photoshopping a dick onto Trump's face is 100% protected expression. Producing a photoreal deepfake of him balls deep in Lindsey Graham's ass while Mitch McConnell can be seen holding the camera in a mirror wearing a ballgag and cuck strap then posting it online either without context or trying to pass it off as real is a problem.

Adalast OP ,

This could be a good option. I will have to look into it. I know some of our family is not the most savvy (lucky to be able to use FB) so I may have to look into building a front end on top of it for them, but this is a solid start.

Adalast OP ,

I will happily look at the alternatives. We avoid Apple like it carries the plague, mostly on my objections to their licensing policies alone. Also, I love that you linked to something about The Fappening, have a 💯 and my heartiest appreciation for you as a scholar and a gentleman.

Adalast OP ,

This has potential in many ways. I will have to set it up and see how it feels.

Adalast OP ,

Thank you on the congrats! We are so happy and he could not be more adorable.

Also thank you for the suggestion, I will dig into it.

Adalast OP ,

Ideally, as easy to interface with as possible for non-tech literate users. My mother-in-law once told my wife "I don't know why you would ever want to strengthen your mind." in response to confronting my wife on why she was reading a book outside as a child instead of playing physically. This is a mantra she has continued well into her 50's and is still going "strong". I need something she can access and download pictures from to print off and hang on her wall like she does from FB now. This is essentially the low bar. Everyone else should be more competent than that.

Adalast OP ,

I don't mind self-hosting, I just need something that can host from a dynamic IP since static is too expensive right now.

Adalast OP ,

I'm not above getting my hands dirty and this sounds like it could have promise. Thank you.

Adalast OP ,

I will look at QuickConnect as that sounds potentially ideal.

I honestly don't trust MS as far as I could throw them. The amount of ads they are forcing into the OS level is evidence enough for me to believe that they are willing to abuse customers. And if DropBox is any indication of how ToS and EULAs can change in the blink of an eye to include all files, past and present, to be used for AI training with no recourse to opt-out, then MS's current ToS doesn't really give any fuzzy feelings.

I will definitely have to look at dyndns as I need to find a way to provide a static endpoint to gain access to ethically sourced AI training materials for my own works and that sounds like it might work.

And yes, I do work in AI, which is why I am so focused on not allowing the megacorps to ignore even the most basic regimes of ethics or customer respect.

Adalast ,

Big difference is consent.

Adalast ,

I've been planning on doing like a 6 month short on it if I have the money because it is going to be tanked hard.

How to constructively protest against AI voice transcription at work?

As a medical doctor I extensively use digital voice recorders to document my work. My secretary does the transcription. As a cost saving measure the process is soon intended to be replaced by AI-powered transcription, trained on each doctor's voice. As I understand it the model created is not being stored locally and I have no...

Adalast ,

To expound on this, AI models are extremely narrow in scope. One which reproduces audio it is trained on is entirely different from one that understands what is being said. As Mr. Turkalino mentioned, the transcription AIs are built on a combination of speech recognition and incredibly specialized text data that is narrowly defined by your industry (medical in this case). In fact, they may have tuned specific models for separate disciplines. This included thousands of documents ranging from textbooks to scholarly journals along with thousands of recordings of professionals saying the words in a variety of accents and dialects so it can understand the difference between very important and very different sounding words, my wife is pregnant, so amnioitis and amniocentesis come to mind. They are close enough sounding that a general model might mistake them, and that being transcribed wrong could spell real problems when others may look at the patients chart if there are complications.

Also, most models are run in the cloud because the calculations can he very taxing. I run Stable Diffusion and other AIs locally on my beast of a machine and it struggles at times. Realistically, the cloud machines are just bugger than you can get as a desktop. Also, under the most ideal circumstances, the audio of your notes does not live in the servers, it is transmitted, stored on a virtual machine (VM) while it is being processed, then after the results are completed the VM is destroyed and the audio recording goes with it. Nothing is kept. Of course, that is where you need to be sure to do the work, making sure that your situation is "ideal". One of the biggest controversies in with AI right now is that data is being stored for doing reinforcement training on the AI models. Example, you send your recordings and the AI returns the transcript. You mark any corrections and go on with your day. The company takes those recordings and feeds them back into the general model with the corrections you made and tries to tell the AI what it got wrong. You are going to want to be sure that you are allowed to opt-out of your data being allowed to be used as training data (beyond the fine-tuning to help it learn your voice).

Is there a term for being right and then your opposition getting taught a lesson proving you were right? EPA calls off cyber regulations for water sector. (cyberscoop.com)

On July 25, 2023, the states of Missouri, Arkansas, and Iowa, along with intervenors American Water Works Association and National Rural Water Association, petitioned the Eighth Circuit to review the EPA’s new rule. This rule requires states to review and report cybersecurity threats to their public water systems (PWS)....

Adalast ,

So what I'm hearing is that any cyber "vindication" should be targeted at the highest income communities in the states. Gotcha.

Adalast ,

Yeah, I initially was ok with it, but as I have watched these companies I have become less and less ok. I have been contemplating making dummy accounts full of erroneous data so all of the metrics are wrong as a giant middle finger. Sure, I'm a 72-year-old woman in Des Moines, or am I an 80-year-old man in DC? Maybe a 22-year-old in LA? Who knows.

Adalast ,

If you own it, don't install a damn thing your employer demands. If they want security access on a device, they pay for it.

If you don't own it, don't use it for a damn thing that isn't work-related and use it minimally for that. "Yes sir" emails and submitting reports. That's it. Don't do research, don't surf the web, don't accept a single personal call, email, or text. They don't have any right to know anything that is not work related.

Adalast ,

Ok, but so do most humans? So few people actually have true understanding in topics. They parrot the parroting that they have been told throughout their lives. This only gets worse as you move into more technical topics. Ask someone why it is cold in winter and you will be lucky if they say it is because the days are shorter than in summer. That is the most rudimentary "correct" way to answer that question and it is still an incorrect parroting of something they have been told.

Ask yourself, what do you actually understand? How many topics could you be asked "why?" on repeatedly and actually be able to answer more than 4 or 5 times. I know I have a few. I also know what I am not able to do that with.

Adalast ,

I am so glad I had one that was the opposite. I discussed practical applications of the subject material after class with him and at the end of the semester he gave me a B+ even though I only got a C by score because I actually grasped the material better than anyone else in the class, even if I was not able to evaluate it as well on the tests.

Adalast ,

In some ways, you are correct. It is coming though. The psychological/neurological word you are searching for is "conceptualization". The AI models lack the ability to abstract the text they know into the abstract ideas of the objects, at least in the same way humans do. Technically the ability to say "show me a chair" and it returns images of a chair, then following up with "show me things related to the last thing you showed me" and it shows couches, butts, tables, etc. is a conceptual abstraction of a sort. The issue comes when you ask "why are those things related to the first thing?" It is coming, but it will be a little while before it is able to describe the abstraction it just did, but it is capable of the first stage at least.

Adalast ,

Lol @ driving a car being simple. That is one of the more complex sensory somatic tasks that humans do. You have to calculate the rate of all vehicles in front of you, assess for collision probabilities, monitor for non-vehicle obstructions (like people, animals, etc.), adjust the accelerator to maintain your own velocity while terrain changes, be alert to any functional changes in your vehicle and be ready to adapt to them, maintain a running inventory of laws which apply to you at the given time and be sure to follow them. Hell, that is not even an exhaustive list for a sunny day under the best conditions. Driving is fucking complicated. We have all just formed strong and deeply connected pathways in our somatosensory and motor cortexes to automate most of the tasks. You might say it is a very well-trained neural network with hundreds to thousands of hours spent refining and perfecting the responses.

The issue that AI has right now is that we are only running 1 to 3 sub-AIs to optimize and calculate results. Once that number goes up, they will be capable of a lot more. For instance: one AI for finding similarities, one for categorizing them, one for mapping them into a use case hierarchy to determine when certain use cases apply, one to analyze structure, one to apply human kineodynamics to the structure and a final one to analyze for effectiveness of the kineodynamic use cases when done by a human. This would be a structure that could be presented an object and told that humans use it and the AI brain could be able to piece together possible uses for the tool and describe them back to the presenter with instructions on how to do so.

Adalast ,

I had a shower fantasy the other day where I wrote an AI that hacked all of the data brokers and web trackers then doxxed every politician and political candidate's web history in it's totality the instant they were added to the ballot, worldwide. It was decentralized running many instances and used a similar structure for distributing the information.

Adalast ,

Name 5 that did not have sweeping adverse consequences, with accompanying sources. I will even accept Wikipedia pages if they have attributions. Make sure they are major ones that really shaped the course of human existence moving forward from their introduction.

Adalast ,

Don't forget a healthy dose of Dunning-Kruger suppositories.

Adalast ,

"key features" being their spyware bot that they only got working in Windows and Mac.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • incremental_games
  • meta
  • All magazines
    •