sunday Posted January 8 Share Posted January 8 Now that is progress! Link to comment Share on other sites More sharing options...
lucklucky Posted January 8 Author Share Posted January 8 Progress in fakery. Link to comment Share on other sites More sharing options...
lucklucky Posted January 12 Author Share Posted January 12 (edited) https://www.notebookcheck.net/Popular-tech-site-CNET-reportedly-employing-AI-to-write-full-fledged-articles.680974.0.html Quote With the launch of OpenAI’s ChatGPT in November 2022, many thought that it was only a matter of time before AI-generated articles flood the Search Engine Result Pages (SERP). Google maintains that content produced using machine learning tools is categorized as spam which, in theory, is penalized by the search giant. As it turns out, CNET, one of the biggest tech media sites in the world, has been publishing articles generated with AI since November 2022. According to Gael Breton of Authority Hacker, CNET has been putting up articles written with AI since November 11, 2022. To date, the site has 72 such pieces. While CNET does disclose under each piece that the article was written using “automation technology”, it is not immediately clear, as the byline only mentions “Written by CNET Money”. It is only after clicking on the byline that you are presented with the disclaimer that the content was generated with an AI engine. Edited January 12 by lucklucky Link to comment Share on other sites More sharing options...
DB Posted January 12 Share Posted January 12 Fairly sure I've been seeing Facebook recommendations for pseudo-articles on military matters. They read like poorer quality Indian blog posts but don't even have a hint of factual content, always outrageous futurist nonsense with photshops of absurd hardware. Link to comment Share on other sites More sharing options...
sunday Posted January 12 Share Posted January 12 On 1/8/2023 at 9:23 PM, lucklucky said: Progress in fakery. Well, yes, but there are legit applications. I've found a bit unsettling to videoconferencing with guys that are not looking at my eyes, but at my image in the screen, because of the parallax difference with the camera lens. 43 minutes ago, lucklucky said: https://www.notebookcheck.net/Popular-tech-site-CNET-reportedly-employing-AI-to-write-full-fledged-articles.680974.0.html Seems CNET could not afford a Stuart. Link to comment Share on other sites More sharing options...
lucklucky Posted January 12 Author Share Posted January 12 Hahah Link to comment Share on other sites More sharing options...
sunday Posted January 13 Share Posted January 13 American Students Cheat On Tests Using ChatGPT, Professor Marks Them 'Fail' Quote (...) To check his suspicions, the professor plugged the essay into the software made by the makers of ChatGPT, and the software showed a 99.9 percent match with what a computer program would write. Darren even tried ChatGPT on his own and fed it questions that the students might have entered and got a similar essay as that of his students who cheated, though he did not get any identical answers since the program is built to produce somewhat unique answers. Later, when the student was confronted, he confessed to using the software and was marked "fail" in the subject. Link to comment Share on other sites More sharing options...
DB Posted January 13 Share Posted January 13 So now add "cheater" to "bad debtor" to his digital identity. I doubt that McDonalds would employ him. Link to comment Share on other sites More sharing options...
lucklucky Posted January 14 Author Share Posted January 14 U.S. Blacklists YMTC, 21 Chinese Companies on AI Threat https://www.eetimes.com/u-s-blacklists-ymtc-21-chinese-companies-on-ai-threat/ Quote The U.S. Commerce Department is blacklisting Yangtze Memory Technologies Co. (YMTC) and more than 20 other Chinese chipmakers suspected of an AI threat to U.S. national security. The Commerce Department’s Bureau of Industry and Security (BIS) expanded measures taken in October by adding YMTC and companies like AI chip designer Cambricon Technologies to its so-called Entity List, blocking their ability to buy semiconductors and production tools that are made with U.S. technology. “Today, we are building on the actions we took in October to protect U.S. national security by severely restricting the PRC’s ability to leverage artificial intelligence, advanced computing, and other powerful, commercially available technologies for military modernization and human rights abuses,” Alan Estevez, Under Secretary of Commerce for Industry and Security, said in prepared remarks issued Wednesday. “This work will continue.” Link to comment Share on other sites More sharing options...
lucklucky Posted January 17 Author Share Posted January 17 Link to comment Share on other sites More sharing options...
DB Posted January 18 Share Posted January 18 This might kill Stable Diffusion (if Getty wins, I think they're done), and have a significant impact on any future use of image generators that use a training set formed of copyrighted imagery. However, the cork is out of the bottle and you can't put it back in. https://www.bbc.co.uk/news/technology-64285227 Link to comment Share on other sites More sharing options...
lucklucky Posted January 19 Author Share Posted January 19 How can Getty goes after freeware? Link to comment Share on other sites More sharing options...
DB Posted January 19 Share Posted January 19 So, the point is that the company behind it has scraped millions of images off the internet without respecting the copyright clauses associated with them. Their use goes far beyond fair use by any reasonable definition and the only potential defence is likely to be the "value added" one of being a derivative work (as I understand it, but copyright law varies across countries and I don't even pretend to know it well.) So, if Getty can demonstrate that the use of these scraped images violates copyright, then they're done. If they've gained commercially from it, they're very done. Link to comment Share on other sites More sharing options...
lucklucky Posted January 20 Author Share Posted January 20 Stable diffusion exists even in freeware tools. Link to comment Share on other sites More sharing options...
Sardaukar Posted January 20 Share Posted January 20 We need Butlerian Jihad... Link to comment Share on other sites More sharing options...
TrustMe Posted January 20 Share Posted January 20 43 minutes ago, Sardaukar said: We need Butlerian Jihad... If we did that how will TankNet survive Link to comment Share on other sites More sharing options...
lucklucky Posted January 22 Author Share Posted January 22 (edited) Not sure if true or mockup(so fake) but shows how are we heading in creation https://blenderartists.org/t/new-technologies/1341536/227 Edited January 22 by lucklucky Link to comment Share on other sites More sharing options...
Burncycle360 Posted January 23 Share Posted January 23 (edited) Best time to invest was yesterday, next best time is today. The various AI are reaching critical mass. What's next is the wild west, like the dot com rush and crypto, and what will emerge over the next 10 years will be every bit as destabilizing as the internet and social media... everyone is scrambling to figure out how to make money with this, from teenagers in basements to google itself holding emergency meetings. Old legislators don't even have an awareness understanding of what this means ("the internet is a series of tubes!") and so are not functionally capable of implementing wise regulation on par with the founding fathers. Using copyright material to teach it as a legal claim may be a speed bump but the genie is out of the bottle, and frankly I don't think it will hold water because it will be the equivalent of no expectation of privacy when in public. You can't stop some designer from being inspired by a sweet car he sees on the road and bringing fins back. Just looking is not a crime, and just like OSINT if the public can see it from a google image search, the argument is that AI can too. It's just faster, and our existing paradigm wasn't designed to handle that. Edited January 23 by Burncycle360 Link to comment Share on other sites More sharing options...
Burncycle360 Posted January 23 Share Posted January 23 It won't be long until the future amazons of the industry start to set themselves apart, integrating the capabilities of the various AI. Deepfakes will tap into Voice AI and ChatGPT like natural language AI to start producing content that's difficult to tell from real. Traditional corporate media is still struggling to adapt to the internet age to this day, and they're way behind the curve with regards to staying relevant in the age of social media, this on top of that is their death knell, something new is going to emerge. The sheer volume of articles on par with new york times all the way down to vice and tabloids that can be produced basically instantly is just unfathomable. A few engineers can do the job of an entire staff of hundreds of woke employees, or the right wing equivalent. The media content the average person is exposed to is already not reflective of real life (as in the views presented are not proportional to those who hold them in the population), being mostly captured by activist leftists. Unfortunately it still has an effect on the sentiment of the layperson, which translates into votes for populist demagogues and ultimately policy and laws. AI exacerbates this problem for a variety of reasons. You can generate orders of magnitude more content in a fraction of the time. It was already difficult to inoculate kids against this phenomenon with critical thinking skills, and it's going to be nearly impossible now. When this capability clashes with precedents regarding the free market, regulation and government limitations on overreach, we will have no way to reign it in without violating some of those precedents set on principle. Link to comment Share on other sites More sharing options...
lucklucky Posted January 24 Author Share Posted January 24 (edited) I think this might favour the right that is not that inclined to evangelisation and seems today to have an uncanny incapability to tell and write the stories. The cultural output of the right is much lower compared to the left. I agree this will change the world as we know. It has the potential to be much more disruptive than internet. Edited January 24 by lucklucky Link to comment Share on other sites More sharing options...
lucklucky Posted January 24 Author Share Posted January 24 https://www.latimes.com/entertainment-arts/business/story/2022-12-19/the-next-frontier-in-moviemaking-ai-edits Link to comment Share on other sites More sharing options...
DB Posted January 24 Share Posted January 24 Now that's interesting - I think that lip-syncing for translations falls within my personal threshold for acceptable use, and maybe the word replacement to get a lower rating is ok (although there is an artistic argument that messing with dialog like this is poor form, the censorship is the price of doing business). Leaving aside CGI use for the superhero leaps and invulnerability, the whole of the rest of it is, well, at what point do you just admit it's photo-realistic animé? Link to comment Share on other sites More sharing options...
Soren Ras Posted January 25 Share Posted January 25 A Butlerian Jihad is looking more attractive every day… Link to comment Share on other sites More sharing options...
lucklucky Posted January 25 Author Share Posted January 25 Advanced AI 'could kill everyone’, warn Oxford researchers MPs told technology has become a ‘literal arms race’ for countries and area should be regulated like nuclear weapons https://www.telegraph.co.uk/news/2023/01/25/advanced-ai-could-kill-everyone-warn-oxford-researchers/ Link to comment Share on other sites More sharing options...
Burncycle360 Posted January 25 Share Posted January 25 Imagine something on par with the cuban missile crisis but with decision makers on both sides being fed AI generated content blasts from all media sources sufficient in volume to obfuscate what is actually happening... Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now