Will Schenk
@willschenk.com
father, widower, lover, technologist, woodsman
We have different opinions. This is not the end of the world.
November 11, 2025 at 12:32 PM
We have different opinions. This is not the end of the world.
Which is mainly that their whole business model is built on top of stealing other peoples things while funneling money away from publishers. LLMs just refactor how search works, the newspaper and media industries are still in the same sinking ship as they were before.
November 11, 2025 at 1:58 AM
Which is mainly that their whole business model is built on top of stealing other peoples things while funneling money away from publishers. LLMs just refactor how search works, the newspaper and media industries are still in the same sinking ship as they were before.
This is just rephrasing what I said -- that you don't see an ethical issue with google search but you do with using that to train an LLM. I do see an ethical issue with google search, and I see the LLM being the exact same ethical issue, no more. So that's where we diverge in our opinions.
November 11, 2025 at 1:57 AM
This is just rephrasing what I said -- that you don't see an ethical issue with google search but you do with using that to train an LLM. I do see an ethical issue with google search, and I see the LLM being the exact same ethical issue, no more. So that's where we diverge in our opinions.
I'm trying to understand your position. You don't seem to think that google steals the internet data to run their business, and that LLM training is somehow different. Cool.
November 11, 2025 at 12:39 AM
I'm trying to understand your position. You don't seem to think that google steals the internet data to run their business, and that LLM training is somehow different. Cool.
I've tried quite hard to get it to return copyright material, and that's simply not true. But ok, that's the feeling. You probably haven't run tests. The google crawlers have been sucking in copyrighted material without permission since the beginning. They got into trouble with google books.
November 11, 2025 at 12:31 AM
I've tried quite hard to get it to return copyright material, and that's simply not true. But ok, that's the feeling. You probably haven't run tests. The google crawlers have been sucking in copyrighted material without permission since the beginning. They got into trouble with google books.
anyway I'm just trying to figure out the point of view, I'm not trying to convince anyone if anything
November 10, 2025 at 11:31 PM
anyway I'm just trying to figure out the point of view, I'm not trying to convince anyone if anything
I honestly don't understand what you are referring to. Do you think that the llm will spit out copyright material?
November 10, 2025 at 11:04 PM
I honestly don't understand what you are referring to. Do you think that the llm will spit out copyright material?
LLMs are quite literally the extension of the Google auto complete. Same engineers who built that were involved with the Attention is all you need paper. Same lab built the first LLM model. I mean that they are literally the same, not as a metaphor or whatever.
November 9, 2025 at 7:23 PM
LLMs are quite literally the extension of the Google auto complete. Same engineers who built that were involved with the Attention is all you need paper. Same lab built the first LLM model. I mean that they are literally the same, not as a metaphor or whatever.
Google quite literally sucks up all the worlds information. It's very literal not paid for, not sure if that counts as stolen. All the search training is on that data.
November 9, 2025 at 7:20 PM
Google quite literally sucks up all the worlds information. It's very literal not paid for, not sure if that counts as stolen. All the search training is on that data.
I'm the CTO of an AI engineering company, so I do know that there isn't a hard difference between them. They are all simply predicting next tokens. Same way that your Google search results are tuned by what other people click on. I'm curious how you see it though.
November 9, 2025 at 7:12 PM
I'm the CTO of an AI engineering company, so I do know that there isn't a hard difference between them. They are all simply predicting next tokens. Same way that your Google search results are tuned by what other people click on. I'm curious how you see it though.
What is this thinking behind this? Do you use spell check? Do you use Google? What do you think is the difference?
November 9, 2025 at 2:42 PM
What is this thinking behind this? Do you use spell check? Do you use Google? What do you think is the difference?
You take a picture of it with your banks phone app and it gets deposited into your account. At least, that's what I do, but my bank is in dollars.
June 3, 2025 at 5:15 PM
You take a picture of it with your banks phone app and it gets deposited into your account. At least, that's what I do, but my bank is in dollars.
There's a great firewall of China to keep the western influence out, not the other way around. I'm just saying that it's a weird narrative to push that there aren't American products with American ideology baked into everything
April 19, 2025 at 12:00 AM
There's a great firewall of China to keep the western influence out, not the other way around. I'm just saying that it's a weird narrative to push that there aren't American products with American ideology baked into everything
They were probably posting it on their smartphone over the Internet, most likely on an US company's website. In English.
April 18, 2025 at 11:42 PM
They were probably posting it on their smartphone over the Internet, most likely on an US company's website. In English.
I hope what's written in there still works!
February 28, 2025 at 9:26 AM
I hope what's written in there still works!