Michael Dexter
boosted
In big news overnight, #Anthropic have made a major change to their user data retention and training policy - giving customers until September 28th to opt out, or have their chats, code sessions and other artefacts used for training for up to five years.
This is a major departure from their previous privacy-first stance.
But what's really behind this change? As Connie Loizos points out in this @Techcrunch article, it's all about the #data.
As I've spoken about recently, we've passed #PeakToken - the point in history where we have the maximum amount of authentic, human-generated data available. Now, the internet is polluted with synthetically-generated #AIslop. If you're an #AI company scraping the web for new data to train on, that's bad news, because you also scoop up the AI slop. If models are trained on AI slop, they're likely to encounter #ModelCollapse - like a bad photocopy.
Anthropic's play here is all about the #TokenCrisis - the voracious appetite for new, authentic, human-generated data to train on - part of a broader phenomenon I've termed the #TokenWars.
As new data becomes scarcer and more valuable, it will be more sought after and contested. We're still in the early days of the #TokenWars, and we should expect to see more moves like this to secure more data for AI training.
@KathyReid I dare to predict that this ultra-wasteful industry will be remembered as a sophomoric project.
There is less understanding now of the nature of human reasoning than what scholars had before GPT training.
People who believed that intelligence is in form, the normal, that semantics can be turned into syntax, a marginal, were the derided few Carnap hold-outs. It's still not possible, it still leads to Orwellian absurdity and decay.
Relax and smile.
In big news overnight, #Anthropic have made a major change to their user data retention and training policy - giving customers until September 28th to opt out, or have their chats, code sessions and other artefacts used for training for up to five years.
This is a major departure from their previous privacy-first stance.
But what's really behind this change? As Connie Loizos points out in this @Techcrunch article, it's all about the #data.
As I've spoken about recently, we've passed #PeakToken - the point in history where we have the maximum amount of authentic, human-generated data available. Now, the internet is polluted with synthetically-generated #AIslop. If you're an #AI company scraping the web for new data to train on, that's bad news, because you also scoop up the AI slop. If models are trained on AI slop, they're likely to encounter #ModelCollapse - like a bad photocopy.
Anthropic's play here is all about the #TokenCrisis - the voracious appetite for new, authentic, human-generated data to train on - part of a broader phenomenon I've termed the #TokenWars.
As new data becomes scarcer and more valuable, it will be more sought after and contested. We're still in the early days of the #TokenWars, and we should expect to see more moves like this to secure more data for AI training.
Angela Antunovic
and 8 others
boosted
Hacker News front page today.
Uh-oh.
"AWS CEO says using AI to replace junior staff is 'Dumbest thing I've ever heard'"
https://www.theregister.com/2025/08/21/aws_ceo_entry_level_jobs_opinion/
"Mark Zuckerberg freezes AI hiring amid bubble fears"
https://www.telegraph.co.uk/business/2025/08/21/zuckerberg-freezes-ai-hiring-amid-bubble-fears/
"95% of Companies See ‘Zero Return’ on $30 Billion Generative AI Spend, MIT Report Finds"
https://thedailyadda.com/95-of-companies-see-zero-return-on-30-billion-generative-ai-spend-mit-report-finds/
Hacker News front page today.
Uh-oh.
"AWS CEO says using AI to replace junior staff is 'Dumbest thing I've ever heard'"
https://www.theregister.com/2025/08/21/aws_ceo_entry_level_jobs_opinion/
"Mark Zuckerberg freezes AI hiring amid bubble fears"
https://www.telegraph.co.uk/business/2025/08/21/zuckerberg-freezes-ai-hiring-amid-bubble-fears/
"95% of Companies See ‘Zero Return’ on $30 Billion Generative AI Spend, MIT Report Finds"
https://thedailyadda.com/95-of-companies-see-zero-return-on-30-billion-generative-ai-spend-mit-report-finds/
"I don’t write these posts for VC-funded LLMs to come along and gobble up and produce some shitty facsimile, or summarise what I’m saying with none of the nuance or context on someone else’s website."
Same.
@piccalilli This blog is the CUTEST damn thing.
I am a sucker for pixel art and I agree.
As a writer it really gets on my nerves when I see about 70% of the advertising on YouTube is #aislop now having both generated script and visuals.
hamish campbell
boosted
"A web that you don’t understand is one that’s not user-driven, or driven by user utility. It’s one where people don’t have agency over the technology they depend upon, and where changes aren’t contrived to address a specific user need.
At the risk of sounding as conspiratorial as the author of the original Dead Internet Theory, I believe the emergence of this incomprehensibility was borne of deliberate decisions made by people at the very top of Big Tech. Furthermore, I believe these people are profoundly anti-person, and see people as resources to be tapped rather than collaborators within a vast, global digital ecosystem.
In many respects, I think this phenomena is down to two things: first, many of the tech products we use were founded by people who were still in the throes of youth, and became billionaires and global tech icons before their brains were even fully developed. They’ve been insulated from people from a young age, never lived a normal life, and they’ve been told — repeatedly — that they are geniuses and visionaries. While I don’t think this explanation excuses any of their behavior, I also think it goes some way into explaining the scarcely-disguised antipathy these people show for their fellow humans.
(...)
The other factor behind this phenomenon is that many of the people running these companies are former management consultants spawned from hell (read: McKinsey) and then set loose on what amounts to “essential infrastructure” for the digital age. Sundar Pichai, the CEO of Google, is a former McKinsey consultant. Sheryl Sandberg, the former Chief Operating Officer of Facebook was one, too.
Management consultants have one job — it’s to recommend strategies to cowardly CEOs that they probably thought of themselves, but are too chicken to put their name behind themselves, that invariably screw over employees, consumers, and the planet."
https://whatwelost.substack.com/p/how-the-internet-died
#DeadInternet#Commons#DigitalCommons#OpenWeb#SocialMedia#AISlop#AI
"A web that you don’t understand is one that’s not user-driven, or driven by user utility. It’s one where people don’t have agency over the technology they depend upon, and where changes aren’t contrived to address a specific user need.
At the risk of sounding as conspiratorial as the author of the original Dead Internet Theory, I believe the emergence of this incomprehensibility was borne of deliberate decisions made by people at the very top of Big Tech. Furthermore, I believe these people are profoundly anti-person, and see people as resources to be tapped rather than collaborators within a vast, global digital ecosystem.
In many respects, I think this phenomena is down to two things: first, many of the tech products we use were founded by people who were still in the throes of youth, and became billionaires and global tech icons before their brains were even fully developed. They’ve been insulated from people from a young age, never lived a normal life, and they’ve been told — repeatedly — that they are geniuses and visionaries. While I don’t think this explanation excuses any of their behavior, I also think it goes some way into explaining the scarcely-disguised antipathy these people show for their fellow humans.
(...)
The other factor behind this phenomenon is that many of the people running these companies are former management consultants spawned from hell (read: McKinsey) and then set loose on what amounts to “essential infrastructure” for the digital age. Sundar Pichai, the CEO of Google, is a former McKinsey consultant. Sheryl Sandberg, the former Chief Operating Officer of Facebook was one, too.
Management consultants have one job — it’s to recommend strategies to cowardly CEOs that they probably thought of themselves, but are too chicken to put their name behind themselves, that invariably screw over employees, consumers, and the planet."
https://whatwelost.substack.com/p/how-the-internet-died
#DeadInternet#Commons#DigitalCommons#OpenWeb#SocialMedia#AISlop#AI
Michael Bacon
boosted
> Researchers warn that our increasing use of AI could be limiting our language and thought processes https://www.cbc.ca/news/ai-human-thought-processes-1.7604789
Ah yes, the wonderful #tech that folks in places like #academia and #HigherEd, including #librarians who should know better, are rushing to embrace.
> Researchers warn that our increasing use of AI could be limiting our language and thought processes https://www.cbc.ca/news/ai-human-thought-processes-1.7604789
Ah yes, the wonderful #tech that folks in places like #academia and #HigherEd, including #librarians who should know better, are rushing to embrace.
beka valentine
boosted
