@seanjmay.bsky.social
I mean, they already have this one.
It was the Mr. Hanky / Fantasia / Robert Redford / Sundance episode.
That said, if they were willing to make a sequel to that, I could go for a #2.
It was the Mr. Hanky / Fantasia / Robert Redford / Sundance episode.
That said, if they were willing to make a sequel to that, I could go for a #2.
October 20, 2025 at 12:03 PM
I mean, they already have this one.
It was the Mr. Hanky / Fantasia / Robert Redford / Sundance episode.
That said, if they were willing to make a sequel to that, I could go for a #2.
It was the Mr. Hanky / Fantasia / Robert Redford / Sundance episode.
That said, if they were willing to make a sequel to that, I could go for a #2.
The concern being that in a system where commerce is still what it currently is, that if there is 0 protection for the author (not the IP Dragon’s hoard), that getting paid for making it, when there are 18 carbon copies with different author names and/or titles will just be luck / advertising reach.
October 20, 2025 at 11:41 AM
The concern being that in a system where commerce is still what it currently is, that if there is 0 protection for the author (not the IP Dragon’s hoard), that getting paid for making it, when there are 18 carbon copies with different author names and/or titles will just be luck / advertising reach.
These laws were meant to ensure that an artist would get credit, and would get paid for their creation, for a while, before it became a common good. They were specifically preventing people with no talent, but with access to publishing tools, from saying "that's mine, now; I made it".
...
...
October 20, 2025 at 11:41 AM
These laws were meant to ensure that an artist would get credit, and would get paid for their creation, for a while, before it became a common good. They were specifically preventing people with no talent, but with access to publishing tools, from saying "that's mine, now; I made it".
...
...
Back in the days of Don Quixote, copyright was envisioned as a way to prevent people from slapping their name on your book, and reselling it as their work. Or taking your book, ripping some pages out, and calling it transformative.
This still happens (and I’m not talking about "style").
This still happens (and I’m not talking about "style").
October 20, 2025 at 11:41 AM
Back in the days of Don Quixote, copyright was envisioned as a way to prevent people from slapping their name on your book, and reselling it as their work. Or taking your book, ripping some pages out, and calling it transformative.
This still happens (and I’m not talking about "style").
This still happens (and I’m not talking about "style").
This just in:
Deep Purple wrote a rock song about a filking mishap, which was summarily filked.
Deep Purple wrote a rock song about a filking mishap, which was summarily filked.
September 27, 2025 at 9:24 PM
This just in:
Deep Purple wrote a rock song about a filking mishap, which was summarily filked.
Deep Purple wrote a rock song about a filking mishap, which was summarily filked.
"AI" that all of the tech-bros hype up, will not live up to the hype. Again, all of the hype is to get tax breaks and investments for their LLM.
Other research is getting defunded. That makes it harder to improve, not easier. So it's autocomplete or bust, until the bubble goes.
Other research is getting defunded. That makes it harder to improve, not easier. So it's autocomplete or bust, until the bubble goes.
September 27, 2025 at 4:25 AM
"AI" that all of the tech-bros hype up, will not live up to the hype. Again, all of the hype is to get tax breaks and investments for their LLM.
Other research is getting defunded. That makes it harder to improve, not easier. So it's autocomplete or bust, until the bubble goes.
Other research is getting defunded. That makes it harder to improve, not easier. So it's autocomplete or bust, until the bubble goes.
... le sigh.
Like your LLM of choice suggested, not all "AI" is created equal. Right now, most money that is publicly visible is going into LLMs.
All of the less public money is going into weapons of various kinds.
Is your point that AI will make regular lives better, or be a Musk payout?
Like your LLM of choice suggested, not all "AI" is created equal. Right now, most money that is publicly visible is going into LLMs.
All of the less public money is going into weapons of various kinds.
Is your point that AI will make regular lives better, or be a Musk payout?
September 27, 2025 at 4:04 AM
... le sigh.
Like your LLM of choice suggested, not all "AI" is created equal. Right now, most money that is publicly visible is going into LLMs.
All of the less public money is going into weapons of various kinds.
Is your point that AI will make regular lives better, or be a Musk payout?
Like your LLM of choice suggested, not all "AI" is created equal. Right now, most money that is publicly visible is going into LLMs.
All of the less public money is going into weapons of various kinds.
Is your point that AI will make regular lives better, or be a Musk payout?
I didn't purport to be intellectually superior. Your arguments, to the point where I stepped in were:
"AI will be the next industrial revolution"
and
"As evidenced by GDP"
a great amount of money *could* be going to research for decent usecases. Most is hype, instead. How do you square that?
"AI will be the next industrial revolution"
and
"As evidenced by GDP"
a great amount of money *could* be going to research for decent usecases. Most is hype, instead. How do you square that?
September 27, 2025 at 3:46 AM
I didn't purport to be intellectually superior. Your arguments, to the point where I stepped in were:
"AI will be the next industrial revolution"
and
"As evidenced by GDP"
a great amount of money *could* be going to research for decent usecases. Most is hype, instead. How do you square that?
"AI will be the next industrial revolution"
and
"As evidenced by GDP"
a great amount of money *could* be going to research for decent usecases. Most is hype, instead. How do you square that?
Again, put it into the LLM that gave you your definition of "AI"
September 27, 2025 at 3:16 AM
Again, put it into the LLM that gave you your definition of "AI"
Right. If you don't have a definition for an em-dash on hand, nor an opinion on what CMoS, nor Strunk & White, say about them, then I would hazard to presume your definition of AI was, indeed, definitionally not yours.
September 27, 2025 at 2:31 AM
Right. If you don't have a definition for an em-dash on hand, nor an opinion on what CMoS, nor Strunk & White, say about them, then I would hazard to presume your definition of AI was, indeed, definitionally not yours.
Well, you can get those definitions from the same place you got the em-dash.
If you have examples of people with technical knowledge saying that LLMs are suddenly going to... not be autocomplete... who aren't invested in companies providing companies offering LLMs, I’d sure like to hear them.
If you have examples of people with technical knowledge saying that LLMs are suddenly going to... not be autocomplete... who aren't invested in companies providing companies offering LLMs, I’d sure like to hear them.
September 27, 2025 at 2:04 AM
Well, you can get those definitions from the same place you got the em-dash.
If you have examples of people with technical knowledge saying that LLMs are suddenly going to... not be autocomplete... who aren't invested in companies providing companies offering LLMs, I’d sure like to hear them.
If you have examples of people with technical knowledge saying that LLMs are suddenly going to... not be autocomplete... who aren't invested in companies providing companies offering LLMs, I’d sure like to hear them.
Is that an em?
Again, if you think that the technology that the VC-bros are all about is the same technology that futurists are talking about, you are as incorrect as the paper that was penned by all of the LLM heads, demanding money to prevent vectorized-autocomplete from becoming Skynet.
Again, if you think that the technology that the VC-bros are all about is the same technology that futurists are talking about, you are as incorrect as the paper that was penned by all of the LLM heads, demanding money to prevent vectorized-autocomplete from becoming Skynet.
September 27, 2025 at 1:45 AM
Is that an em?
Again, if you think that the technology that the VC-bros are all about is the same technology that futurists are talking about, you are as incorrect as the paper that was penned by all of the LLM heads, demanding money to prevent vectorized-autocomplete from becoming Skynet.
Again, if you think that the technology that the VC-bros are all about is the same technology that futurists are talking about, you are as incorrect as the paper that was penned by all of the LLM heads, demanding money to prevent vectorized-autocomplete from becoming Skynet.
Given I am referring to the actual types of technology, and you are using "ai" and talking about how it will advance, while also suggesting that "ai" has seen massive GDP growth, thus conflating an *awful lot of things*, you should first define "AI".
September 27, 2025 at 12:15 AM
Given I am referring to the actual types of technology, and you are using "ai" and talking about how it will advance, while also suggesting that "ai" has seen massive GDP growth, thus conflating an *awful lot of things*, you should first define "AI".
I wrote my first Perceptron, more than a decade ago, during the last AI winter.
I wrote CNNs while people were looking at style transfer.
I didn't say "AI", I said: "show me the papers that say LLMs (VC-bro AI) will achieve sentience, let alone super-intelligence".
I wrote CNNs while people were looking at style transfer.
I didn't say "AI", I said: "show me the papers that say LLMs (VC-bro AI) will achieve sentience, let alone super-intelligence".
September 27, 2025 at 12:09 AM
I wrote my first Perceptron, more than a decade ago, during the last AI winter.
I wrote CNNs while people were looking at style transfer.
I didn't say "AI", I said: "show me the papers that say LLMs (VC-bro AI) will achieve sentience, let alone super-intelligence".
I wrote CNNs while people were looking at style transfer.
I didn't say "AI", I said: "show me the papers that say LLMs (VC-bro AI) will achieve sentience, let alone super-intelligence".
Well, your argument did nothing to account for the distribution of the GDP increase, so why not the single point of Altman's wallet?
The difference of our two statements has 0 bearing on GDP, which is your target metric.
The difference of our two statements has 0 bearing on GDP, which is your target metric.
September 26, 2025 at 11:53 PM
Well, your argument did nothing to account for the distribution of the GDP increase, so why not the single point of Altman's wallet?
The difference of our two statements has 0 bearing on GDP, which is your target metric.
The difference of our two statements has 0 bearing on GDP, which is your target metric.
Your argument is that if we give infinite money specifically to Sam Altman, that GDP will rise, infinitely.
September 26, 2025 at 1:36 PM
Your argument is that if we give infinite money specifically to Sam Altman, that GDP will rise, infinitely.
Again, please explain how an ML neural network, which is not an LLM, has anything to do with the LLMs of all of the companies making claims, currently, about how their LLMs are going to attain sentience and become the singularity (OpenAI et al), and so they need billions in tax breaks and grants.
September 25, 2025 at 8:08 PM
Again, please explain how an ML neural network, which is not an LLM, has anything to do with the LLMs of all of the companies making claims, currently, about how their LLMs are going to attain sentience and become the singularity (OpenAI et al), and so they need billions in tax breaks and grants.
Please provide research papers on how Large Language Models (and spdcifically LLMs) will achieve sentience.
My keyboard autocorrect has become much worse, since it became LLM-driven.
Most disruption has more to do with managers cutting heads, rather than actual LLM capabilities.
My keyboard autocorrect has become much worse, since it became LLM-driven.
Most disruption has more to do with managers cutting heads, rather than actual LLM capabilities.
September 25, 2025 at 7:01 PM
Please provide research papers on how Large Language Models (and spdcifically LLMs) will achieve sentience.
My keyboard autocorrect has become much worse, since it became LLM-driven.
Most disruption has more to do with managers cutting heads, rather than actual LLM capabilities.
My keyboard autocorrect has become much worse, since it became LLM-driven.
Most disruption has more to do with managers cutting heads, rather than actual LLM capabilities.
Ionian
DeLorean
Nightly raid of the Phrygian
but my life is mostly Locrian
DeLorean
Nightly raid of the Phrygian
but my life is mostly Locrian
September 3, 2025 at 8:38 PM
Ionian
DeLorean
Nightly raid of the Phrygian
but my life is mostly Locrian
DeLorean
Nightly raid of the Phrygian
but my life is mostly Locrian
But I have been at places where the newbies are onboarded into a bunch of comments, and then their first contributions completely ignore the body of a function, because they interpreted something else from the comment. And this is back when you had to hallucinate your own errors, with no outsourcing
September 3, 2025 at 1:53 AM
But I have been at places where the newbies are onboarded into a bunch of comments, and then their first contributions completely ignore the body of a function, because they interpreted something else from the comment. And this is back when you had to hallucinate your own errors, with no outsourcing
And my personal belief is that if a whole team is familiar with *a pattern* and that pattern is actually self-documenting, either via human language, or via something like category theory (and there's no mutation, and it's declarative and pure and referentially transparent) then 0 is great
September 3, 2025 at 1:53 AM
And my personal belief is that if a whole team is familiar with *a pattern* and that pattern is actually self-documenting, either via human language, or via something like category theory (and there's no mutation, and it's declarative and pure and referentially transparent) then 0 is great
I generally tell greener developers to ignore the comments, in established ("legacy") code, if the comment is telling you what it is doing, instead of why.
If you just take the comment's word for it, and it screws up, the actual instructions were right there.
I’ve definitely known "lone wolves"
If you just take the comment's word for it, and it screws up, the actual instructions were right there.
I’ve definitely known "lone wolves"
September 3, 2025 at 1:53 AM
I generally tell greener developers to ignore the comments, in established ("legacy") code, if the comment is telling you what it is doing, instead of why.
If you just take the comment's word for it, and it screws up, the actual instructions were right there.
I’ve definitely known "lone wolves"
If you just take the comment's word for it, and it screws up, the actual instructions were right there.
I’ve definitely known "lone wolves"
Sure. Do you know if it was even a government employee?
Can the government be held responsible for the actions of an unidentified citizen? Who is going to take on that case? Who will the supreme court side with, when the constitutionality of the ruling is questioned?
Do try. But expect troubles.
Can the government be held responsible for the actions of an unidentified citizen? Who is going to take on that case? Who will the supreme court side with, when the constitutionality of the ruling is questioned?
Do try. But expect troubles.
August 29, 2025 at 9:38 PM
Sure. Do you know if it was even a government employee?
Can the government be held responsible for the actions of an unidentified citizen? Who is going to take on that case? Who will the supreme court side with, when the constitutionality of the ruling is questioned?
Do try. But expect troubles.
Can the government be held responsible for the actions of an unidentified citizen? Who is going to take on that case? Who will the supreme court side with, when the constitutionality of the ruling is questioned?
Do try. But expect troubles.
...sue... whom?
If a person in a full cartoon villain outfit, including balaclava, broke in and kidnapped your kid, how would you know who to sue, exactly?
If a person in a full cartoon villain outfit, including balaclava, broke in and kidnapped your kid, how would you know who to sue, exactly?
August 29, 2025 at 1:27 PM
...sue... whom?
If a person in a full cartoon villain outfit, including balaclava, broke in and kidnapped your kid, how would you know who to sue, exactly?
If a person in a full cartoon villain outfit, including balaclava, broke in and kidnapped your kid, how would you know who to sue, exactly?
Wasn't she the one who tried to tell the ladies how to work a Nine-Tooth Hive?
August 22, 2025 at 9:09 AM
Wasn't she the one who tried to tell the ladies how to work a Nine-Tooth Hive?