I think that this is the scariest part of the upcoming AI era.
Every time you want to do something, you essentially have to ask permission from a corporation (and, possibly, from a government that regulates it) to do it.
I was thinking about Bitcoin and Satoshi Nakamoto.
Regardless of what you think about Bitcoin (I know that large part of HN crowd is very critical of it), my point is that people were able to create software that has big social impact without asking permission from anyone and without revealing their identity.
Back in the days you could be just some guy sitting in your room, possibly in a backwaters 3rd world country, with a laptop that costs just a couple of hundred dollars and you (potentially) could be just as impactful as a well funded corporate coder. It all depended on your skills and ingenuity.
Now, we are coming to an era when you:
1. Have to have resources to buy tokens or buy/rent inference hardware
2. Have to reveal your identity, your intentions and ask for a permission. Everything you do is potentially logged and analyzed
Of course, theoretically you can do everything "the old school way", but if AI makes a programmer 100x - 1000x more productive, you will always be behind and never catch up.
It's a difference between peer-to-peer Bitcoin decentralization and extreme AI datacenter centralization. Open weight models currently also require centralized datacenters for training, even though inference can be done decentralized.. We will see whether centralization or decentralization wins long term.
Every closed source software has the potential to centralize and lock you out of it. This has been true before and will be true after the AI bubble (not saying AI isn't useful but the current valuations will pop at some point, just like dotcom).
Bitcoin was decentralized but so was cash. And it was still possible to send cash to people around the world it was just harder to set up and trust. And those cash sending mechanisms didn't really lock people out of them (but could theoretically, sometimes did by way of sanctions).
AI doesn't make programmers 100x productive and it never will, so not really a point there. As long as the output is non deterministic, you have to check it, and no one can check their code 100x faster than writing it
> Every closed source software has the potential to centralize and lock you out of it.
Ye, but you could (at least, theoretically) make your own open source alternative and you weren't particularly disadvantaged compared to corporate programmers doing closed source.
For example, Linux initially was just one guy doing it for fun and see how far did it go.
We may soon come to a situation, when something like Linux will be impossible, because devs making proprietary software will have access to newer/better models that are only sold to big companies and tightly controlled (in the sense that you can use them only for certain "safe" purposes).
> Bitcoin was decentralized but so was cash. And it was still possible to send cash to people around the world it was just harder to set up and trust. And those cash sending mechanisms didn't really lock people out of them (but could theoretically, sometimes did by way of sanctions).
My point is that in future we may come to a scenario when someone like Satoshi won't be able to create something like Bitcoin while keeping their privacy. Because you have to provide a phone number to use OpenAI and also because "creating a non government-authorized financial instrument violates my safety guidelines" or something.
> AI doesn't make programmers 100x productive and it never will, so not really a point there. As long as the output is non deterministic, you have to check it, and no one can check their code 100x faster than writing it
Maybe, maybe not. I'm afraid that eventually fully automated systems will be able to produce code that is good enough without manual checking. Doesn't matter if it is deterministic or not. If the probability of a catastrophic failure is low enough, you can safely ignore it.
It’s not just correcting grammar. Not just a few words. Almost every paragraph starts in the same style.
To be clear, I don’t want to be negative. Don’t want to just troll —- I’m just noticing how all of it is generated. It’s difficult to unsee once you know.
The article could have benefited from a mention of why Anthropic is like this now.
The reason this happened to you is Anthropic's systems presumably think your work looks like (Chinese) distillation and subscription-to-token-reseller-api practices.
But that's a reasonable guess at best, and knowing my own usage _nothing_ would have made my activity look like that. I was not maxing out usage limits, I was not running things 24/7, and my activity peaked on my typical working hours.
And that's exactly the type of situation where I wish Anthropic would have just put me in front of a human to review OR at least let me know what in my usage caused flags to be raised so I can avoid them in the future
No idea. My initial guess was because I was running long-running Fable 5 tasks related to healthcare, and I think that's within a specific classification of usage, but then why wouldn't they have suspended me earlier -- it doesn't make sense
In competitive online games, anticheat bans come in waves to make them more difficult to reverse-engineer. Clearly Anthropic is being deliberately opaque here. I wonder if you tripped some anti-distillation measures?
Yeah that's our leading theory too -- but I'm wracking my brain trying to think about exactly what in my activity caused it. It's not like I was exhausting my usage limits nor working 24/7 with random-ish corpora, it's been pretty typical desk work usage. :shrug:
A lot of GEPA-inspired prompt tuning with content healthcare related, which I assumed may have been what raised some flags on their end -- but I've been going at this for a long time so I don't get why they suspended _now_
What got me was "I keep seeing the same shape next to billing events in other people's reports".
Clicked off after that.
I'm extremely bullish on AI, but I am tired of people not using their own words. Is everyone so insecure about the personality they project in writing that they need to replace it with some AI bullshit?
(To the author: if you didn't do this, which I accept is a possibility, I'm sorry -- you've been caught by the storm of fatigue that has plagued us due to those who do replace their words with AI crap)
Hmm. You are making a good point, and I wouldn’t give the benefit of the doubt to this post’s author. But you can’t just unsee “this is twice load bearing”. Some people will wonder if it’s a training, self-reinforcing feedback artifact. For others, it will work like a memetic virus[^1].
The paradoxical thing is that those of us who have English as a second language should have an easier time producing non-AI-English by the mere act of directly writing what we want to write.
And? I’d rather reader poorly written English written by a person about his actual experience - or just read nothing at all - as opposed to reading thinking and writing that has been outsourced to a machine. Being able to think and write confers privileges on those who can do both; the idea that somebody should be able to benefit by doing neither is pretty fucking perverse.
I think that this is the scariest part of the upcoming AI era.
Every time you want to do something, you essentially have to ask permission from a corporation (and, possibly, from a government that regulates it) to do it.
I was thinking about Bitcoin and Satoshi Nakamoto.
Regardless of what you think about Bitcoin (I know that large part of HN crowd is very critical of it), my point is that people were able to create software that has big social impact without asking permission from anyone and without revealing their identity.
Back in the days you could be just some guy sitting in your room, possibly in a backwaters 3rd world country, with a laptop that costs just a couple of hundred dollars and you (potentially) could be just as impactful as a well funded corporate coder. It all depended on your skills and ingenuity.
Now, we are coming to an era when you:
1. Have to have resources to buy tokens or buy/rent inference hardware
2. Have to reveal your identity, your intentions and ask for a permission. Everything you do is potentially logged and analyzed
Of course, theoretically you can do everything "the old school way", but if AI makes a programmer 100x - 1000x more productive, you will always be behind and never catch up.
It's a difference between peer-to-peer Bitcoin decentralization and extreme AI datacenter centralization. Open weight models currently also require centralized datacenters for training, even though inference can be done decentralized.. We will see whether centralization or decentralization wins long term.
Every closed source software has the potential to centralize and lock you out of it. This has been true before and will be true after the AI bubble (not saying AI isn't useful but the current valuations will pop at some point, just like dotcom).
Bitcoin was decentralized but so was cash. And it was still possible to send cash to people around the world it was just harder to set up and trust. And those cash sending mechanisms didn't really lock people out of them (but could theoretically, sometimes did by way of sanctions).
AI doesn't make programmers 100x productive and it never will, so not really a point there. As long as the output is non deterministic, you have to check it, and no one can check their code 100x faster than writing it
> Every closed source software has the potential to centralize and lock you out of it.
Ye, but you could (at least, theoretically) make your own open source alternative and you weren't particularly disadvantaged compared to corporate programmers doing closed source.
For example, Linux initially was just one guy doing it for fun and see how far did it go.
We may soon come to a situation, when something like Linux will be impossible, because devs making proprietary software will have access to newer/better models that are only sold to big companies and tightly controlled (in the sense that you can use them only for certain "safe" purposes).
> Bitcoin was decentralized but so was cash. And it was still possible to send cash to people around the world it was just harder to set up and trust. And those cash sending mechanisms didn't really lock people out of them (but could theoretically, sometimes did by way of sanctions).
My point is that in future we may come to a scenario when someone like Satoshi won't be able to create something like Bitcoin while keeping their privacy. Because you have to provide a phone number to use OpenAI and also because "creating a non government-authorized financial instrument violates my safety guidelines" or something.
> AI doesn't make programmers 100x productive and it never will, so not really a point there. As long as the output is non deterministic, you have to check it, and no one can check their code 100x faster than writing it
Maybe, maybe not. I'm afraid that eventually fully automated systems will be able to produce code that is good enough without manual checking. Doesn't matter if it is deterministic or not. If the probability of a catastrophic failure is low enough, you can safely ignore it.
>Not a warning. Not a rate limit. A revocation.
Clearly OP has some sock puppet claude accounts that weren't banned
I still have Grok around to correct my grammar
It’s not just correcting grammar. Not just a few words. Almost every paragraph starts in the same style.
To be clear, I don’t want to be negative. Don’t want to just troll —- I’m just noticing how all of it is generated. It’s difficult to unsee once you know.
Your response sure looks AI-generated.
You're absolutely right!
(At our office, we jokingly use AI-isms as we accuse each other of being meat proxies for our AI overlords.)
The article could have benefited from a mention of why Anthropic is like this now. The reason this happened to you is Anthropic's systems presumably think your work looks like (Chinese) distillation and subscription-to-token-reseller-api practices.
But that's a reasonable guess at best, and knowing my own usage _nothing_ would have made my activity look like that. I was not maxing out usage limits, I was not running things 24/7, and my activity peaked on my typical working hours.
And that's exactly the type of situation where I wish Anthropic would have just put me in front of a human to review OR at least let me know what in my usage caused flags to be raised so I can avoid them in the future
Oh whoops I ceded my livelihood to an opaque corporate subscription service and they banned me
You can probably blame Chinese distillation efforts for some of this paranoia.
So you have no idea why they banned you? None at all?
No idea. My initial guess was because I was running long-running Fable 5 tasks related to healthcare, and I think that's within a specific classification of usage, but then why wouldn't they have suspended me earlier -- it doesn't make sense
In competitive online games, anticheat bans come in waves to make them more difficult to reverse-engineer. Clearly Anthropic is being deliberately opaque here. I wonder if you tripped some anti-distillation measures?
Yeah that's our leading theory too -- but I'm wracking my brain trying to think about exactly what in my activity caused it. It's not like I was exhausting my usage limits nor working 24/7 with random-ish corpora, it's been pretty typical desk work usage. :shrug:
What kind of work were you doing?
A lot of GEPA-inspired prompt tuning with content healthcare related, which I assumed may have been what raised some flags on their end -- but I've been going at this for a long time so I don't get why they suspended _now_
You're absolutely right. Same product family. Same sudden wall.
This is load bearing
What got me was "I keep seeing the same shape next to billing events in other people's reports".
Clicked off after that.
I'm extremely bullish on AI, but I am tired of people not using their own words. Is everyone so insecure about the personality they project in writing that they need to replace it with some AI bullshit?
(To the author: if you didn't do this, which I accept is a possibility, I'm sorry -- you've been caught by the storm of fatigue that has plagued us due to those who do replace their words with AI crap)
Hmm. You are making a good point, and I wouldn’t give the benefit of the doubt to this post’s author. But you can’t just unsee “this is twice load bearing”. Some people will wonder if it’s a training, self-reinforcing feedback artifact. For others, it will work like a memetic virus[^1].
The paradoxical thing is that those of us who have English as a second language should have an easier time producing non-AI-English by the mere act of directly writing what we want to write.
[^1]: https://en.wikipedia.org/wiki/Viruses_of_the_Mind
Have you thought about the fact that the author is in the Philippines and English isn’t his first language?
And? I’d rather reader poorly written English written by a person about his actual experience - or just read nothing at all - as opposed to reading thinking and writing that has been outsourced to a machine. Being able to think and write confers privileges on those who can do both; the idea that somebody should be able to benefit by doing neither is pretty fucking perverse.
At least write it yourself. Why would anybody trust an AI’s narrative of events?