#enshittfication — Public Fediverse posts

One week after deleting my LLM accounts

I deleted my ChatGPT and Claude accounts last week. It’s vanishingly unlikely this will be a permanent state of affairs but I did think it was a useful exercise to prepare myself for a future of enshittified chatbots. These systems are going to get more expensive to use and less reliable in the short to medium term. Therefore I’d like to establish where they are actually adding value to my existence and where I have fallen into using them out of habit. A few observations after week one:

• There are interesting topics which occur to me that I default to exploring with a chatbot. In some cases I use other means to explore this but mostly I just let them drop. For example I found myself wondering yesterday about the economics of escape rooms which I think would have previously led to a half hour conversation with ChatGPT or Claude. The level of interest just isn’t high enough to lead me to explore this without immediate access to a chatbot to satisfy my curiosity.
• The first time I had a practical question I slipped into a conversation with Google Search before I realised “hey wait I’m supposed to not be doing this”. This illustrates how ubiquitous models have become in our digital lives. It also made me realise that I do think for many practical order questions it really is a more effective means of finding good enough answers than searching. I’m fairly certain ChatGPT could have told me how to turn off the faulty ‘replace filter’ light in my air filter in one shot, whereas I still haven’t got round to looking up the instruction book to get the manufacturer’s advice on troubleshooting.
• There was a health and fitness question I would have defaulted to discussing with a chatbot. In the absence of it I found myself thinking “hey wait I know this stuff” and actually thinking it through myself based on my existing knowledge. This was the only sign I saw this week of a kind of outsourcing that troubled me where defaulting to a chatbot was something I was deriving confidence from rather than using a purely neutral tool.
• There are practical machinic tasks which I’ve started to rely on them for. I thought I might re-organise my blog at some point before I restart work (as the kind of thing I’d intended to do for months but never got round to with usual routines). But then I remembered I don’t have a Claude account anymore. The WordPress MCP is so powerful I just couldn’t imagine doing this kind of tedious work directly any more even when the outcome matters to me.
• I feel notably less digitally distracted. I’m off work and not on e-mail so that’s almost certainly the major factor but I suspect this is a meaningful contributing factor nonetheless. I spend a lot of time each week reading text and writing text through interacting with chatbots so the intensivity and extent of my screentime is significantly reduced.
• I’m surprised how strong my instinct is to share this blog post with a chatbot while writing it. I suddenly realise how much I’d slipped into sharing pretty much everything I write with Claude Opus. Not because I want its input on the content but rather because it so reliably offers “here are some more things to think about on this theme”. I also don’t think I’d realised how frequently I was using it as a light touch copy editor. The stream of consciousness way in which I write blog posts feels ever so slightly exposing to me now in a way it didn’t.

My plan is to wait until August at least before setting up new accounts. Ideally I’ll hold off until the absence of models is materially impacting on my life in an obviously negative way. I suspect there’s some deeper psychological aspects to daily use of models which haven’t really shown up in my experience yet, which might after a few weeks or months.

One week after deleting my LLM accounts

I deleted my ChatGPT and Claude accounts last week. It’s vanishingly unlikely this will be a permanent state of affairs but I did think it was a useful exercise to prepare myself for a future of enshittified chatbots. These systems are going to get more expensive to use and less reliable in the short to medium term. Therefore I’d like to establish where they are actually adding value to my existence and where I have fallen into using them out of habit. A few observations after week one:

• There are interesting topics which occur to me that I default to exploring with a chatbot. In some cases I use other means to explore this but mostly I just let them drop. For example I found myself wondering yesterday about the economics of escape rooms which I think would have previously led to a half hour conversation with ChatGPT or Claude. The level of interest just isn’t high enough to lead me to explore this without immediate access to a chatbot to satisfy my curiosity.
• The first time I had a practical question I slipped into a conversation with Google Search before I realised “hey wait I’m supposed to not be doing this”. This illustrates how ubiquitous models have become in our digital lives. It also made me realise that I do think for many practical order questions it really is a more effective means of finding good enough answers than searching. I’m fairly certain ChatGPT could have told me how to turn off the faulty ‘replace filter’ light in my air filter in one shot, whereas I still haven’t got round to looking up the instruction book to get the manufacturer’s advice on troubleshooting.
• There was a health and fitness question I would have defaulted to discussing with a chatbot. In the absence of it I found myself thinking “hey wait I know this stuff” and actually thinking it through myself based on my existing knowledge. This was the only sign I saw this week of a kind of outsourcing that troubled me where defaulting to a chatbot was something I was deriving confidence from rather than using a purely neutral tool.
• There are practical machinic tasks which I’ve started to rely on them for. I thought I might re-organise my blog at some point before I restart work (as the kind of thing I’d intended to do for months but never got round to with usual routines). But then I remembered I don’t have a Claude account anymore. The WordPress MCP is so powerful I just couldn’t imagine doing this kind of tedious work directly any more even when the outcome matters to me.
• I feel notably less digitally distracted. I’m off work and not on e-mail so that’s almost certainly the major factor but I suspect this is a meaningful contributing factor nonetheless. I spend a lot of time each week reading text and writing text through interacting with chatbots so the intensivity and extent of my screentime is significantly reduced.
• I’m surprised how strong my instinct is to share this blog post with a chatbot while writing it. I suddenly realise how much I’d slipped into sharing pretty much everything I write with Claude Opus. Not because I want its input on the content but rather because it so reliably offers “here are some more things to think about on this theme”. I also don’t think I’d realised how frequently I was using it as a light touch copy editor. The stream of consciousness way in which I write blog posts feels ever so slightly exposing to me now in a way it didn’t.

My plan is to wait until August at least before setting up new accounts. Ideally I’ll hold off until the absence of models is materially impacting on my life in an obviously negative way. I suspect there’s some deeper psychological aspects to daily use of models which haven’t really shown up in my experience yet, which might after a few weeks or months.

One week after deleting my LLM accounts

I deleted my ChatGPT and Claude accounts last week. It’s vanishingly unlikely this will be a permanent state of affairs but I did think it was a useful exercise to prepare myself for a future of enshittified chatbots. These systems are going to get more expensive to use and less reliable in the short to medium term. Therefore I’d like to establish where they are actually adding value to my existence and where I have fallen into using them out of habit. A few observations after week one:

• There are interesting topics which occur to me that I default to exploring with a chatbot. In some cases I use other means to explore this but mostly I just let them drop. For example I found myself wondering yesterday about the economics of escape rooms which I think would have previously led to a half hour conversation with ChatGPT or Claude. The level of interest just isn’t high enough to lead me to explore this without immediate access to a chatbot to satisfy my curiosity.
• The first time I had a practical question I slipped into a conversation with Google Search before I realised “hey wait I’m supposed to not be doing this”. This illustrates how ubiquitous models have become in our digital lives. It also made me realise that I do think for many practical order questions it really is a more effective means of finding good enough answers than searching. I’m fairly certain ChatGPT could have told me how to turn off the faulty ‘replace filter’ light in my air filter in one shot, whereas I still haven’t got round to looking up the instruction book to get the manufacturer’s advice on troubleshooting.
• There was a health and fitness question I would have defaulted to discussing with a chatbot. In the absence of it I found myself thinking “hey wait I know this stuff” and actually thinking it through myself based on my existing knowledge. This was the only sign I saw this week of a kind of outsourcing that troubled me where defaulting to a chatbot was something I was deriving confidence from rather than using a purely neutral tool.
• There are practical machinic tasks which I’ve started to rely on them for. I thought I might re-organise my blog at some point before I restart work (as the kind of thing I’d intended to do for months but never got round to with usual routines). But then I remembered I don’t have a Claude account anymore. The WordPress MCP is so powerful I just couldn’t imagine doing this kind of tedious work directly any more even when the outcome matters to me.
• I feel notably less digitally distracted. I’m off work and not on e-mail so that’s almost certainly the major factor but I suspect this is a meaningful contributing factor nonetheless. I spend a lot of time each week reading text and writing text through interacting with chatbots so the intensivity and extent of my screentime is significantly reduced.
• I’m surprised how strong my instinct is to share this blog post with a chatbot while writing it. I suddenly realise how much I’d slipped into sharing pretty much everything I write with Claude Opus. Not because I want its input on the content but rather because it so reliably offers “here are some more things to think about on this theme”. I also don’t think I’d realised how frequently I was using it as a light touch copy editor. The stream of consciousness way in which I write blog posts feels ever so slightly exposing to me now in a way it didn’t.

My plan is to wait until August at least before setting up new accounts. Ideally I’ll hold off until the absence of models is materially impacting on my life in an obviously negative way. I suspect there’s some deeper psychological aspects to daily use of models which haven’t really shown up in my experience yet, which might after a few weeks or months.

Microsoft's May 2026 upgrade fail just cost us $250 in data recovery and many many hours of lost time trying to restore and reconfigure my wife's laptop.

The impunity with which these large corporations can screw over their "customers" (read: hostages) is despicable.

#microsoft #enshittfication

Microsoft's May 2026 upgrade fail just cost us $250 in data recovery and many many hours of lost time trying to restore and reconfigure my wife's laptop.

The impunity with which these large corporations can screw over their "customers" (read: hostages) is despicable.

#microsoft #enshittfication

@lxplm
Das Prinzip #enshittfication von @pluralistic wurde hier schön übersetzt ins analoge Leben. Und am Schluss die Feststellung: "Du bildest es dir nicht ein. Digitale Dienste werden schlechter. Aber erfreulicherweise muss das nicht so sein." 💟

Mehr Info und den Bericht "Breaking free – pathways to a fair technological future" gibt es hier.
https://www.forbrukerradet.no/breakingfree

@lxplm
Das Prinzip #enshittfication von @pluralistic wurde hier schön übersetzt ins analoge Leben. Und am Schluss die Feststellung: "Du bildest es dir nicht ein. Digitale Dienste werden schlechter. Aber erfreulicherweise muss das nicht so sein." 💟

Mehr Info und den Bericht "Breaking free – pathways to a fair technological future" gibt es hier.
https://www.forbrukerradet.no/breakingfree

#Netzpolitik:
"
Stirb langsam: Warum es mit X nun zu Ende geht

Soziale Netzwerke sind wegen des Netzwerkeffekts nur schwer totzukriegen. Doch dem Twitter-Nachfolger X von Elon Musk droht nun genau das. ..
"
https://netzpolitik.org/2024/stirb-langsam-warum-es-mit-x-nun-zu-ende-geht/

14.12.2024

#Ausbeutung #BlueSky #Desinformation #Enshittfication #Hass #InterestGraph #Kommerzialisierung #Machtmissbrauch #Musk #Netzwerkeffekt #Propaganda #SocialMedia #Twitter #Verschwörungstheorie #X #Zensur

#Netzpolitik:
"
Stirb langsam: Warum es mit X nun zu Ende geht

Soziale Netzwerke sind wegen des Netzwerkeffekts nur schwer totzukriegen. Doch dem Twitter-Nachfolger X von Elon Musk droht nun genau das. ..
"
https://netzpolitik.org/2024/stirb-langsam-warum-es-mit-x-nun-zu-ende-geht/

14.12.2024

#Ausbeutung #BlueSky #Desinformation #Enshittfication #Hass #InterestGraph #Kommerzialisierung #Machtmissbrauch #Musk #Netzwerkeffekt #Propaganda #SocialMedia #Twitter #Verschwörungstheorie #X #Zensur

Today we talk about #enshittfication, yet the univeral phenomena of software spoilage has been a topic much longer. Such as in this great 2007 post from Jeff Atwood: https://blog.codinghorror.com/why-does-software-spoil/

Today we talk about #enshittfication, yet the univeral phenomena of software spoilage has been a topic much longer. Such as in this great 2007 post from Jeff Atwood: https://blog.codinghorror.com/why-does-software-spoil/

Olha só! O dicionário australiano Macquarie escolheu a palavra criada pelo @pluralistic, #enshittfication, como palavra do ano de 2024!

Vi numa rede que o Cory incluiu nas fileiras da #merdificação 😅
(Valeu Bruno Murtinho!)

https://www.aljazeera.com/news/2024/11/26/what-many-of-us-feel-enshittification-named-word-of-the-year

Olha só! O dicionário australiano Macquarie escolheu a palavra criada pelo @pluralistic, #enshittfication, como palavra do ano de 2024!

Vi numa rede que o Cory incluiu nas fileiras da #merdificação 😅
(Valeu Bruno Murtinho!)

https://www.aljazeera.com/news/2024/11/26/what-many-of-us-feel-enshittification-named-word-of-the-year

Just dusted off my newsletter to publish a proposal for the glut of consumer AI we're seeing now, with a simple proposal:

Build as much AI as you would like into your products, but do so in a way that allows users to bring their own AIs, instead of being forced to use the platforms':

https://www.notesfromthevoid.cc/p/consumer-ai-a-simple-proposal

#GenerativeAI #Enshittfication #LockIn #ConsumerChoice #UserControl #Antitrust #ArtificialIntelligence #LargeLanguageModels #Android #GoogleIO

Just dusted off my newsletter to publish a proposal for the glut of consumer AI we're seeing now, with a simple proposal:

Build as much AI as you would like into your products, but do so in a way that allows users to bring their own AIs, instead of being forced to use the platforms':

https://www.notesfromthevoid.cc/p/consumer-ai-a-simple-proposal

#GenerativeAI #Enshittfication #LockIn #ConsumerChoice #UserControl #Antitrust #ArtificialIntelligence #LargeLanguageModels #Android #GoogleIO

I forgot that to Denuvo, switching proton versions looks like you're installing it on a bunch of new Windows computers.

Why do I need Denuvo in a game that's only single player again? And one that Steam verifies that I purchased.

#enshittfication

Looks like I need to move away from Authy. I still need it for work but using it for personal 2FA seems like a losing proposition. #enshittfication

Looks like I need to move away from Authy. I still need it for work but using it for personal 2FA seems like a losing proposition. #enshittfication

If dishwashers were invented today

🔵 you would have to buy special packets only usable by that brand of washer
🔵 the washer would have to maintain a connection to the internet
🔵 only approved brand dishes could be washed - they would have a special ‘wash chip’ in them to enable
🔵 you would have to purchase a wash subscription service to allow the washer to run longer than 10 mins
🔵 the washer would play an ad between every session
🔵 the washer could only be serviced by approved vendors
🔵 every 6 months you would have a special filter that would expire and you have to purchase a new one
🔵 in order to to buy a new house you are required to purchase an approved dish washer
🔵 if you want to wash forks you must purchase a forking upgrade subscription
🔵 you cannot sell the dishwasher to anyone else as the geolocation built in to the device will brick the machine
🔵 if you use unapproved soap the washer will seize up and require an approved technician to reset it for a fee

#enshittfication
#enshittify

If dishwashers were invented today

🔵 you would have to buy special packets only usable by that brand of washer
🔵 the washer would have to maintain a connection to the internet
🔵 only approved brand dishes could be washed - they would have a special ‘wash chip’ in them to enable
🔵 you would have to purchase a wash subscription service to allow the washer to run longer than 10 mins
🔵 the washer would play an ad between every session
🔵 the washer could only be serviced by approved vendors
🔵 every 6 months you would have a special filter that would expire and you have to purchase a new one
🔵 in order to to buy a new house you are required to purchase an approved dish washer
🔵 if you want to wash forks you must purchase a forking upgrade subscription
🔵 you cannot sell the dishwasher to anyone else as the geolocation built in to the device will brick the machine
🔵 if you use unapproved soap the washer will seize up and require an approved technician to reset it for a fee

#enshittfication
#enshittify

@adamshostack @mattblaze it would be nice if it weren't #enshittfication, but, it's google... they'll manage it eventually...

@adamshostack @mattblaze it would be nice if it weren't #enshittfication, but, it's google... they'll manage it eventually...

The #Reddit Protest Is Finally Over. Reddit Won.

Honestly I feel sad that #enshittfication can’t drag user away. If Reddit can move on after these kind of things I think #TwitterX could move on as well. I feel sad that both #Twitter and #Reddit still being go-to place after this kind of #enshittfication….

https://gizmodo.com/reddit-news-blackout-protest-is-finally-over-reddit-won-1850707509
#redditprotest #redditMigration #twittermigration

@Publicenema

Also mir gehts ähnlich und weil mir diese #Enshittfication von #Technologie und mangelnde Optionen andererseits aufn Sack gehen, will ich was besseres sehen...

https://github.com/KBtechnologies/nucbook

@atomicpoet @doctorow 's #enshittfication curve applies.

@atomicpoet @doctorow 's #enshittfication curve applies.