Quick question for anyone who makes it here at all:
What email provider do you use to subscribe here, and did this post arrive in your inbox? Ahh, hell, what a dumb question. Nobody who got this in their junk folder is going to respond.
Three of my four test email providers dumped this post in the "Junk" folders, and the fourth one took 30 minutes to receive it. This is only the second time this has ever happened, and the last time was also when I was critical of Substack. Strange...
Gmail - primary inbox. I am also having deliverability issues, recently cleaned up my list. You just gave me a great idea to post the same comment tomorrow on my stack, so thanks!
Thanks NF. One protonmail address of mine (my actual GC email for this substack!) dumped it in the junk folder. Never happened before! along with a test gmail, and a test gmx. My other proton got it 30 minutes later.
Less than 700 emails received, or less than 5% of my total readership after 2.5 hours. Also never happened before.
Substack's internal 'deliverability' AI detects 'keywords' and lowers post 'reputation score' then sends 'junk' label to email providers if critical of Substack or includes CEO?
This place is turning into a fucking dumpster fire. I'm off to use AI to flood Amazon with those mass market WW2 historical fiction books, the ones that are all just horrible copies of All the light we cannot see, that women can't restrain themselves from gorging on, the ones with disabled children in French villages exploring their love against the backdrop of the greatest indoctrination narrative in history, the ones with lavender flying off the covers designed in Microsoft Paint. Wish me luck.
This is my last year with them. Next summer I'm moving on. Just not sure where. Will seek out something with a server farm buried under an Icelandic volcano that nobody has ever heard of, with a sheep herder holding a double-barreled shotgun guarding the door.
I use my own domains, thirdparadigm.org and retrometro.com. Do I get brownie points for that? It came straight through but had zero comments, making me suspicious something was up. When I refreshed just now, after your 'light n' fluffy' repost intro, it had 78.
That's how I used to get around YT censorship, btw. I found that I could repost the exact video that had been censored with a bland description, new intro and less provocative title than How to Lie With Statistics. I only risked that at the warning level, not when it started going back two years to find some new offense just before my probation was about to expire.
For the rest of my comments on this essay, great inversion: "I love a good inspirational story that is deeply compassionate and human, and can pull on our collective heartstrings, but my mind exists to keep those heartstrings from pulling my chain." Totally agree.
Throughout the essay, I was thinking about Lies are Unbekoming. I had just upped to a paid sub and that seems to be the kiss of death. Except for female Middle Eastern journalists, I seem to always regret it but this is the fastest turnaround I'd seen.
I subbed because a book review was the exact information I was looking for, even having a placeholder title in my book draft, OMGdess. I noticed several books reviewed had feminine topics, and checked the 'about' to see that was a man. My comment ended with 'Are you really one person? How is that possible?'
On the Note version, one of my readers replied that she'd wondered if it was AI generated and clocked some of the repetitions and stylistic tells. But mostly, who has time for that much content?
I usually receive your posts in my junk folder, and I receive emails on my webhosting account email facilities. This time I got nothing. That's why I always keep a tab in my browser open to your home page on here, and check it regularly.
I use Outlook, and your posts normally come into my 'junk' mail, but this morning the cute little ponies arrived into my regular email and I was unable to click 'read full story'--had to come to the Substack site to have a look at what is obviously a formidable piece.
Now that I've read through the whole piece, I've learned a lot. Thank you for your pledge. It's been a while since I've had any interest in most 'output'--what you've written throws some light on why so much of it seems stale. Unlike your writing, which is very human.
Thanks, Paloma. Will reply to your email from last week soon. I was busy with this post and had to exhale yesterday. Can you right-click my posts in your junk folder and select "not junk" on the next few posts to train your client that they should go to the inbox instead?
Nice post, GC, you hit on a lot of relevant points for what I'm also seeing in the digital landscape today. And very nice job calling out Joshua Stylman in particular - every post of his reads strongly AI-assisted (I mentioned his over-reliance briefly in a prior post of mine), and I do recall that he got the COVID deathjabs at the time they came out - and people's true character comes out in times of intense stress, not when times are "good". Thatβs why his posts feel like a continuation of that same pattern of outsourcing judgment rather than wrestling with reality directly. Many other points to make, so I'll run through them briefly:
1. Our upper elites have recognized that they no longer need censorship because they can continue to flood the zone with shit, with endless hot takes, and the vast majority will no longer be able to tell truth from falsehood. This is a good post on the topic: https://blog.exitgroup.us/p/were-all-schizo-posters-now
2. I'm not as allergic to AI use as you seem to be here - there are some writers that reject it entirely (Jasun Horsley, Clintavo), and I respect them on this, but I think LLMs can be used as a tool, although they are quite dangerous from a number of angles (they may manipulate you, feed you false information, make you overly-reliant on them, or they may serve as Narcissus mirrors telling you whatever you want to hear; not to mention their upcoming role in the digital panopticon CBDC/woke AI demiurgic Mark of the Beast system). I like to use it as an editing tool, to provide a brief one paragraph summary of the post, and *sometimes* to assist with research - for example, my post on symbolic speech was so far out there, with so little existing framing in this world, that it took a lot of deep dive conversation with AI to flesh out what my own opinions were on the matter: https://neofeudalreview.substack.com/p/words-as-forcefields-the-exile-of .
3. Every action has a consequence, and for those like Stylman who overuse LLMs, it results in an evisceration of independent thought. So the underlying question is: what is the main purpose of the writer doing the writing? Is it for attention, is it for money? Because if those are the main drivers, then overuse of LLM is no big deal. However, if the main purpose (and there may be other purposes, of course) is spiritual growth, then overuse of LLM *is* a big deal, because it detracts from the actual goal. Real understanding requires sitting with contradiction, with uncertainty, and struggling through the ideas even if it is much slower. Most people havn't done the ugly and sustained introspective work to look into their actual motivations (although we can never know our Self fully even if we try, we can merely approach it).
Like you, I don't want to read LLM-generated output, although I don't mind so much if it's used as an assist. The outsourcing of the basic functionality of writing, though, is a deal breaker to me, and I've unfollowed quite a few writers as a result. But I also don't think it helps to bury one's head in the sand and ignore the technology entirely, either, and it is a process of trial and error to sniff out exactly where that fine line is between the proper use of it as a tool and limiting one's use of it, one I am still navigating.
Oh, lastly: the *way* that people use LLMs is an interesting and under-explored topic, I think. Most people use LLMs for business, for email generation, or for fact based research. To use it as a symbolic tool, an almost oracular function, to investigate deeper meanings behind language itself, seems like a rare use of it, and I don't really hit up against alignment guardrails much as a result. I have a future post prepared on this, and another on McLuhan's tetrad of media effects as applied to LLMs...
Spot on NF. This is the default state of the machines, especially the last point of pleasing the user. An unhappy user will find another machine that lathers them: "...they may manipulate you, feed you false information, make you overly-reliant on them, or they may serve as Narcissus mirrors telling you whatever you want to hear"
I'm not against it for workflows (tech mostly), including extracting text from large volumes, but I always request page numbers to verify the text myself before using it. For research with parameters and furnished books or articles, it's fantastic to save time.
But I started this Substack as a writer, two years before I even opened an LLM, and to many of your most salient points, the more I found myself asking questions or "improving text" in those examples I provided, the more I felt my creative brain rotting, and found in many instances I was more than capable of producing the creative outcome without the damn machine's help. I may have been lazy, but I never needed it. It's a crutch that I will never lean on again for this Substack (text extraction, statistics and math(s)) excluded.
So it's just best for me and my readers going forward that I avoid it in the way I've pledged that I will.
"So the underlying question is: what is the main purpose of the writer doing the writing? Is it for attention, is it for money? Because if those are the main drivers, then overuse of LLM is no big deal." - I disagree with you here. It is a form of deceit and deceit is soul crushing. It dampens the spirit for sure.
Thank you for coming clean on your own use of AI. I hope you have cured yourself of that dependency. My worry is not that the future will be filled with people who think like machines, but people who can't think because they have outsourced that capacity to machines. Maybe that's the same thing. I feel I have become one of those critical readers who can spot the earmarks of AI writing, and when I encounter them on Substack, I stop reading immediately. Partly it's the principle--I reject the use of AI--and partly it is that I do not want to add AI's patterns and tropes to my own thought patterns that come out in my writing, since reading is how people learn how to write. I am interested in almost everything that Lies are Unbekoming writes about, but once I realized he was using AI, I stopped reading him. As you point out, the sheer volume of "content" put out by him and others using AI is another giveaway. I'd also add that many such articles are exceptionally long--a 22-minute read, a 45-minute read, sometimes over an hour. I can't help by feel insulted--and, indeed, infuriated!--by that as a reader. Do these individuals really think they are justified in demanding, or expecting, that degree of time investment from a reader when THEY did not invest their personal time in the writing?
You ended with an excellent question that more and more readers will have to answer as they detect dollar store wikidigest substacks copying and pasting AI text.
Of 200+ posts and 4 years, I used it for polishing or specific vernacular on 12 posts. But this is exactly my thinking why I'll never use it for this Substack nor bother reading past the first paragraph of a ctrl+c/ctrl+v AI substack, it's a highly formulaic brain-rotting machine for the writers out there: "I do not want to add AI's patterns and tropes to my own thought patterns that come out in my writing, since reading is how people learn how to write."
Maybe readers should strike back at those "authors" who are dumping massive amounts of AI text under their byline by using AI summaries to get the gist of those articles. Then we'd have AI writing and AI reading, so the human is 100% left out of the picture! I am kidding, of course. I wouldn't use AI to summarize any more than I'd use it to write. I am an AI tee-totaller! But the scenario is cautionary. How much of our innate capacity to think (reading, writing and thinking are all aspects of the same cognitive function) do we want to lose by deciding not to use them?
I have been diligently striking back! I politely tell every author whose post is clearly written by AI that their post appears to be written by AI. Some take it in stride and confess. Some get huffy and dig their heels in. One author blocked me just for saying that! And she is apparently a truth and freedom loving person who doesnβt censor dissenting opinions on her substack - just embarrassing comments like mine
Good on you! I have said that to one or two authors using AI. One tried to tell me AI can produce wisdom! And this was someone who did not fall for any of the nonsense of the last 5 years and seems to see the danger of digital prison, CBDCs, etc. Yet appears to have fallen into the AI trap.
Wow! Great piece in so many ways! Interesting analysis on Joshua. I subscribed to him liking the "topics" he was addressing. I read little of his work now as I find it "cumbersome". Cumbersome matching the various techniques the AI uses . I guess I demonstrate my naivete in wondering how some of these people could crank out the volume of work they did. I had assumed they, unlike myself, were simply dedicated career writers who did nothing but write. After reading your piece and looking at their "style" I think you pegged it. My background is history and my arena is filled with "court historians" and charlatans that spout agendas for the Crown's coin rather than pursuing the truth. As you have demonstrated, writing is equally contaminated. While I love the substack concept, you hit a major issue on the privacy and anonymity. While the "subscribe" digital currency model is "convenient" and guarantees substack their cut. it creates great vulnerability for the user. Most do not see this, as convenience dominates their thinking. Some of us have experienced the down side first hand, I do not know a solution for mass scale. I know I purchase books of writers I like and have done snail mail contributions to writers that have that option. I suspect most readers are too lazy for that option. To your request, I use Proton to include VPN and proxy server. I get all my articles to my inbox. You do great work and this was truly a needed masterpiece in challenging times.
It's going to be a challenge going forward. As much as the cynic in me says Substack will take the loot and not give a damn, there's a part of me that thinks humans who are not yet accustomed to such obnoxious formulas will soon detect them and get bored with them. Not just the accelerated upscale of 2x "essays" per day, but the machine droning shit writing quality. I dunno, maybe that's naive thinking and they'll prefer shallow pleasures in agreement with Alazons over authenticity and original ideas.
Much like the calculator decimated peopleβs ability to perform basic math, I fear relying on AI to generate content will spell the end of most peopleβs ability to write.
If the current trajectory remains unimpeded, weβre probably only a few generations away from the populace resembling the Eloi, which will make it all the easier for the philanthropathic Morlocks to harvest the herd.
I wonder if neurological studies have been conducted to determine if peopleβs hippocampus has been shrinking due to increasing reliance on smart phones and computers rather than our memories.
Le Bon, Boorstin, Ellul, Postmanβit sounds like youβve been raiding my βFavorite Books on Propaganda, Mass Persuasion, & Psychologyβ (https://margaretannaalice.substack.com/p/favorite-reads) :-) Meerloo would make a good addition to your canon.
βIn a world racing toward mass illiteracy, will humans ever know what Chicken Sands or Orebarger Rings are?β
π€£π€£π€£
βI only use Grammarlyβs free browser add-on for spell-check and nothing more and will contenueu to youtiliz it for yure benifet.β
πππ
Finally, I want to thank you for being honest and honorable about your own use of AI. I have never used it myself and donβt have any intention of doing so, but I must partly thank Emma Sron for warning me away from even trying it as a research tool (https://emmasron.substack.com/p/artificial-intelligence-ai-the-bigger). I donβt want to be complicit in feeding the beast data that may eventually be used to enslave us.
This is a case where I feel like we can use the tools of our enemies to help defeat our enemies as I discussed in a recent post (https://margaretannaalice.substack.com/p/50-reasons-to-give-your-child-video) in which I then quoted from my 2022 Corona Investigative Committee presentation notes, βA Mostly Peaceful Depopulationβ:
P.S. Unbekoming has been so kind to me and appears to be a person of such impeccable character, I almost want to give him a pass for using AI to help with his exhaustive surveys, but I agree disclosure would be the right thing to do when authors do resort to such tools.
For the record, Iβm on ProtonMail and did receive your newsletter.
P.P.S. I do wonder if your email is one that got knocked off my mailing list as I recently attempted to comp you but discovered you werenβt on there. If you want to resubscribe, Iβll be happy to apply a comp.
P.P.P.S. I apologize for the length of this comment, but you can blame your human-composed content for being thought-provoking and engagingβand catching me during a rare window when I wasnβt working on my own content as I was just preparing to crash after my third 20+-hour day in a row.
I am so grateful for this post! A true learning experience.
Having read several of Josh Styleman's postings I would not have had any idea he was using AI generated language. Even though I noticed he was for the most part telling what others (originals such Patrick Wood) have already described in a scholarly way about Technocracy, Styleman's writing flowed easily, and I read it quickly. I get emotional about what is happening to humanity--no longer being human?-- but fooling me is something I must learn to detect more proficiently. You have given me a gift. Thank you! I am not praising you for your admission of using AI in some way in some of your previous work, but it comes across to me like using other tools to get your ideas across in readable, comprehensible form. In the old days, a dictionary or thesaurus or rules on writing a good essay for example (I eventually advanced to spellcheck =:) ). Experimenting with AI to see what it can do is how you are able to teach us who seek knowledge about it. Thank you Professor! Lifting the words whether a phrase to a paragraph of others is abhorrent to me. AI will turn humanity completely against God's Creation as long as it is in the hands of those who's aim is to own the world.
If not using AI means I am an unsophisticated rube of the backwoods, that's fine with me. I'll talk to the trees.
Positively uplifting article, GC. Canβt wait to see all of this AI stuff play out. I want this to just be a stupid fad but only time will tell. Would you find it surprising that use of AI is encouraged by Navy leadership? It is considered a time saving tool.
Thanks, Jennifer. Doesn't surprise me. A perfect addition to other surveillance tools already in use. I suppose this has been the norm for over a decade. AI will probably send any sensitive content warnings directly to supervisors or up the chain. I will say for 'tech' and 'maths' and organizing workflow on projects, it's a game changer, but it all depends on the rewards for the user versus the privacy sacrifices of using it.
Good book, my friend, will have to get back to this one when I have more time!! This one will probably take a couple of Dos Equis to get through. Linking as usual @https://nothingnewunderthesun2016.com/
I use Hotmail and this did get through by the way.
Yeah, I think that maybe the role they're happy to assume now with a higher paid subcount, but also with deteriorating quality. A lot of copy and paste dollar store substacks will be popping up across this platform in the next year, and it will be more difficult to find those unique voices.
You had me at the Killer's song "Human." I usually save myself a lot of time by getting bored with Substack articles before I can finish them which I realized in reading this were probably the AI generated ones.
I also save myself a lot of time by writing all of my own content based on thoughts currently circulating in my head. As I move into deeper spiritual based modes of inquiry lately I'm coming to realize that the only way to reach other humans is to be human...
Recently, I asked Perplexity what I would need to take with me for a trip to Neptune. It gave me a list of things to prepare and have along as if this was an everyday occurrence. I guess it assumed that man has gone to Neptune for a visit. Is the government hiding more crap from us or is A/i as retarded as it can be?
Answer is C. Government is hiding from us just how retarded AI really is. When this bubble pops, it will make pets.com look like a solid investment in 1999.
The pop will likely resemble the words of the illustrious Condaleezza Rice: βWe do not want the evidence of Saddam Husseinβs weaponsβofβmassβdestruction to come in the form of a mushroom cloud.β The mushroom cloud will be the meltdown of nuke plants that are being built to supply power to AI machines. My own sense of things is that that is the point of AI, build something that requires lots of power and use that as the excuse to build nuke plants, then melt them down and depopulate the area except for wildlife ala Chernobyl. I live 5 miles from three mile island which was shut down, I foolishly thought for good until they now are going to get it going again to feed AI and data farms. Data farm being built on the site of a really nice municipally owned golf course. So, as you point out, when the bubble pops, it will really pop. I have to admit I had to use the proton AI "Lumo" to remember Rice's name because I forget everything now, and to get the quote right. Oddly enough when I asked it for her quote about mushroom clouds it gave me this: βWe do not want to see another mushroom cloud over any city, anywhere in the world.β, this quote was from an obscure speech she gave to a bunch of generals and spy types. I had seen her repeat that quote so many times I knew that wasn't it so I had to paraphrase it and then it gave me the right quote. So, did it try to sanitize her infamous quote with one a little less innocuous? I'm very disappointed now because I cannot discern the AI writing you refer to in your article so I will be dependent on yet another stupid machine to figure it out.
This article is one of the reasons I became a paid subscriber, you hit on a lot of things about this world that concern me. Another brilliant piece of writing, who needs AI when people like you still exist. Thankfully I normally realize when a real person is behind the writing and tend to know when someone is no longer producing their own work but has resorted to AI instead. Can't stand it personally, I don't want speed I want writers who have to think about what they are saying and I do want depth.
The rot for me set in long before people had access to computer's and mobile phone's. Back in 1987/88 all parents where I was living at the time received a letter from the schools notifying them they had to provide their children with a calculator for their maths lessons, not a cheap item at the time. No more having to prove the workings of the maths, just tap it out on the calculator, job done. Believe it or not I have only ever used a calculator once in my life and because I couldn't quite trust it I still ended up working it out on paper to check it was right. It was but I realized it just wasn't for me.
By the way, I absolutely loved your " Shameless human-created cyberbegging interludes".
Assuming you are being serious that's terrible, our Good Citizen is the one that stands out, his ability to write intelligently, seriously with that thread of humour running through his writing's and not afraid to say things exactly as they are is worth a gold mine.
Admittedly I had to save to get a yearly subscription but that was more because I dare not do it monthly as here in England the DWP have had access to look through our bank accounts for about a year now to see what we're spending our money on and our government has slowly started chipping away at our retirement pensions in various ways. A monthly payment may have red flagged it as a needless outgoing.
I guess I could have gone for pensioner option but he's worth full payment to me, I'm already saving to upgrade .
Using machines to monetize the consoomer demand for more Content(tm) criticizing the use of machines strikes me as less likely a grasping toward authority and more a reddit smuggie move.
Stupidity is well-distributed across worldviews and ideologies, and the stupid often self-identify by claiming their camp has only smart people (though to be fair some have few or none). So, far from a clever move, monetizing the idiots is dimwitted and self-degrading.
I can think of few admissions-by-action more embarrassing than letting an LLM "improve" one's writing. Maybe a link to an onlyfans account.
Speaking of, don't let LLMs do math for you. They're worse at that than most things they're terrible at. I'll never stop laughing at LLM devs struggling in vain to get massive piles of linear algebra to do basic arithmetic at a millions of times the computational cost of bare metal.
If you need some help with obnoxious math, try WolframAlpha. Its syntax is less tolerant than an LLM but it's a graphing calculator on steroids that runs on almost-plain language and, crucially, actually does the math instead of guessing the answer.
Hahaa. I always need help math. The last course I took was in high school, and it was my only C over four years. I used the upgraded versions of GPT (4 maybe) for statistics calculations on an EV post nearly two years comparing mpg of a Tesla and Honda Accord hybrid ago and didn't check its work. Damn, I hope it wasn't a disaster.
You might also try using AI-assist in spreadsheets. I canβt vouch for their capability (havenβt tried), but at least if the LLM generated a bad formula the calculations would be consistent haha.
Epic Piece, GC. Thank you for all of those carefully-explained AI markers.
We have no doubts about your integrity. Unlike most folks in these times, you know that no one can give to a person his or her Self Respect. People need to 'Take It.....For Themselves.' And, people without self-respect are incapable of respecting anything.
Many decades ago when I was a young lad at Old MacDonald's Farm located at Harmony Junction on Prince Edward Island, local persons who told lies were publicly shunned. Such treatment was highly effective in a time & place with no electricity or home telephones and the nearest "Mountie" was an hour's drive if you had a motor vehicle.
If Samuel Clements were alive in today's JUSA he might write: "Can't swing a dead cat without hitting a paedophile."
Quick question for anyone who makes it here at all:
What email provider do you use to subscribe here, and did this post arrive in your inbox? Ahh, hell, what a dumb question. Nobody who got this in their junk folder is going to respond.
Three of my four test email providers dumped this post in the "Junk" folders, and the fourth one took 30 minutes to receive it. This is only the second time this has ever happened, and the last time was also when I was critical of Substack. Strange...
Gmail and yes.
yahoo sends it to my inbox, however they seem to want to send emails from a board member named "Nuttall" to my spam folder. Perverts
Gmail - primary inbox. I am also having deliverability issues, recently cleaned up my list. You just gave me a great idea to post the same comment tomorrow on my stack, so thanks!
Proton, and yes it arrived.
Thanks NF. One protonmail address of mine (my actual GC email for this substack!) dumped it in the junk folder. Never happened before! along with a test gmail, and a test gmx. My other proton got it 30 minutes later.
Less than 700 emails received, or less than 5% of my total readership after 2.5 hours. Also never happened before.
Substack's internal 'deliverability' AI detects 'keywords' and lowers post 'reputation score' then sends 'junk' label to email providers if critical of Substack or includes CEO?
This place is turning into a fucking dumpster fire. I'm off to use AI to flood Amazon with those mass market WW2 historical fiction books, the ones that are all just horrible copies of All the light we cannot see, that women can't restrain themselves from gorging on, the ones with disabled children in French villages exploring their love against the backdrop of the greatest indoctrination narrative in history, the ones with lavender flying off the covers designed in Microsoft Paint. Wish me luck.
ProtonMail also dumps official ProtonMail communication in the spam folder.
This is my last year with them. Next summer I'm moving on. Just not sure where. Will seek out something with a server farm buried under an Icelandic volcano that nobody has ever heard of, with a sheep herder holding a double-barreled shotgun guarding the door.
Wherever you are and continue to write, I hope I continue to receive your letters.
Machine gun turrets, my dear Citizen. There must be machine gun turrets.
Consider Penang.
Email? or Malaysia?
I got this on Protonmail, but there are surely issues with them now...
I got this on protonmail but other substacks (not yours) have been dumped in proton spam.
Zoho mail. It appears it came straight through.
I just use their App. I've never opened a single newsletter on email, I just delete them (I read deselecting email might unsubscribe so I've avoided).
I use my own domains, thirdparadigm.org and retrometro.com. Do I get brownie points for that? It came straight through but had zero comments, making me suspicious something was up. When I refreshed just now, after your 'light n' fluffy' repost intro, it had 78.
That's how I used to get around YT censorship, btw. I found that I could repost the exact video that had been censored with a bland description, new intro and less provocative title than How to Lie With Statistics. I only risked that at the warning level, not when it started going back two years to find some new offense just before my probation was about to expire.
For the rest of my comments on this essay, great inversion: "I love a good inspirational story that is deeply compassionate and human, and can pull on our collective heartstrings, but my mind exists to keep those heartstrings from pulling my chain." Totally agree.
Throughout the essay, I was thinking about Lies are Unbekoming. I had just upped to a paid sub and that seems to be the kiss of death. Except for female Middle Eastern journalists, I seem to always regret it but this is the fastest turnaround I'd seen.
I subbed because a book review was the exact information I was looking for, even having a placeholder title in my book draft, OMGdess. I noticed several books reviewed had feminine topics, and checked the 'about' to see that was a man. My comment ended with 'Are you really one person? How is that possible?'
On the Note version, one of my readers replied that she'd wondered if it was AI generated and clocked some of the repetitions and stylistic tells. But mostly, who has time for that much content?
I usually receive your posts in my junk folder, and I receive emails on my webhosting account email facilities. This time I got nothing. That's why I always keep a tab in my browser open to your home page on here, and check it regularly.
This is madness. A friend didn't get the post either. I might launch my own writers' coop. Thanks for keeping tabs on me.
protonmail
I received this post at 12:11 PM in my inbox on my Gmail account
proton, and I got it in regular email inbox.
I use Outlook, and your posts normally come into my 'junk' mail, but this morning the cute little ponies arrived into my regular email and I was unable to click 'read full story'--had to come to the Substack site to have a look at what is obviously a formidable piece.
Now that I've read through the whole piece, I've learned a lot. Thank you for your pledge. It's been a while since I've had any interest in most 'output'--what you've written throws some light on why so much of it seems stale. Unlike your writing, which is very human.
Thanks, Paloma. Will reply to your email from last week soon. I was busy with this post and had to exhale yesterday. Can you right-click my posts in your junk folder and select "not junk" on the next few posts to train your client that they should go to the inbox instead?
Gmail and yes it came to my inbox but it does at times send an email I expect to my spam which is why I check spam every other day.
I use pobox.com (now part of fastmail.com) and have it deliver to my own mail server, and it came through just fine about seven hours ago.
Nice post, GC, you hit on a lot of relevant points for what I'm also seeing in the digital landscape today. And very nice job calling out Joshua Stylman in particular - every post of his reads strongly AI-assisted (I mentioned his over-reliance briefly in a prior post of mine), and I do recall that he got the COVID deathjabs at the time they came out - and people's true character comes out in times of intense stress, not when times are "good". Thatβs why his posts feel like a continuation of that same pattern of outsourcing judgment rather than wrestling with reality directly. Many other points to make, so I'll run through them briefly:
1. Our upper elites have recognized that they no longer need censorship because they can continue to flood the zone with shit, with endless hot takes, and the vast majority will no longer be able to tell truth from falsehood. This is a good post on the topic: https://blog.exitgroup.us/p/were-all-schizo-posters-now
2. I'm not as allergic to AI use as you seem to be here - there are some writers that reject it entirely (Jasun Horsley, Clintavo), and I respect them on this, but I think LLMs can be used as a tool, although they are quite dangerous from a number of angles (they may manipulate you, feed you false information, make you overly-reliant on them, or they may serve as Narcissus mirrors telling you whatever you want to hear; not to mention their upcoming role in the digital panopticon CBDC/woke AI demiurgic Mark of the Beast system). I like to use it as an editing tool, to provide a brief one paragraph summary of the post, and *sometimes* to assist with research - for example, my post on symbolic speech was so far out there, with so little existing framing in this world, that it took a lot of deep dive conversation with AI to flesh out what my own opinions were on the matter: https://neofeudalreview.substack.com/p/words-as-forcefields-the-exile-of .
3. Every action has a consequence, and for those like Stylman who overuse LLMs, it results in an evisceration of independent thought. So the underlying question is: what is the main purpose of the writer doing the writing? Is it for attention, is it for money? Because if those are the main drivers, then overuse of LLM is no big deal. However, if the main purpose (and there may be other purposes, of course) is spiritual growth, then overuse of LLM *is* a big deal, because it detracts from the actual goal. Real understanding requires sitting with contradiction, with uncertainty, and struggling through the ideas even if it is much slower. Most people havn't done the ugly and sustained introspective work to look into their actual motivations (although we can never know our Self fully even if we try, we can merely approach it).
Like you, I don't want to read LLM-generated output, although I don't mind so much if it's used as an assist. The outsourcing of the basic functionality of writing, though, is a deal breaker to me, and I've unfollowed quite a few writers as a result. But I also don't think it helps to bury one's head in the sand and ignore the technology entirely, either, and it is a process of trial and error to sniff out exactly where that fine line is between the proper use of it as a tool and limiting one's use of it, one I am still navigating.
Oh, lastly: the *way* that people use LLMs is an interesting and under-explored topic, I think. Most people use LLMs for business, for email generation, or for fact based research. To use it as a symbolic tool, an almost oracular function, to investigate deeper meanings behind language itself, seems like a rare use of it, and I don't really hit up against alignment guardrails much as a result. I have a future post prepared on this, and another on McLuhan's tetrad of media effects as applied to LLMs...
Spot on NF. This is the default state of the machines, especially the last point of pleasing the user. An unhappy user will find another machine that lathers them: "...they may manipulate you, feed you false information, make you overly-reliant on them, or they may serve as Narcissus mirrors telling you whatever you want to hear"
I'm not against it for workflows (tech mostly), including extracting text from large volumes, but I always request page numbers to verify the text myself before using it. For research with parameters and furnished books or articles, it's fantastic to save time.
But I started this Substack as a writer, two years before I even opened an LLM, and to many of your most salient points, the more I found myself asking questions or "improving text" in those examples I provided, the more I felt my creative brain rotting, and found in many instances I was more than capable of producing the creative outcome without the damn machine's help. I may have been lazy, but I never needed it. It's a crutch that I will never lean on again for this Substack (text extraction, statistics and math(s)) excluded.
So it's just best for me and my readers going forward that I avoid it in the way I've pledged that I will.
"So the underlying question is: what is the main purpose of the writer doing the writing? Is it for attention, is it for money? Because if those are the main drivers, then overuse of LLM is no big deal." - I disagree with you here. It is a form of deceit and deceit is soul crushing. It dampens the spirit for sure.
Thank you for coming clean on your own use of AI. I hope you have cured yourself of that dependency. My worry is not that the future will be filled with people who think like machines, but people who can't think because they have outsourced that capacity to machines. Maybe that's the same thing. I feel I have become one of those critical readers who can spot the earmarks of AI writing, and when I encounter them on Substack, I stop reading immediately. Partly it's the principle--I reject the use of AI--and partly it is that I do not want to add AI's patterns and tropes to my own thought patterns that come out in my writing, since reading is how people learn how to write. I am interested in almost everything that Lies are Unbekoming writes about, but once I realized he was using AI, I stopped reading him. As you point out, the sheer volume of "content" put out by him and others using AI is another giveaway. I'd also add that many such articles are exceptionally long--a 22-minute read, a 45-minute read, sometimes over an hour. I can't help by feel insulted--and, indeed, infuriated!--by that as a reader. Do these individuals really think they are justified in demanding, or expecting, that degree of time investment from a reader when THEY did not invest their personal time in the writing?
You ended with an excellent question that more and more readers will have to answer as they detect dollar store wikidigest substacks copying and pasting AI text.
Of 200+ posts and 4 years, I used it for polishing or specific vernacular on 12 posts. But this is exactly my thinking why I'll never use it for this Substack nor bother reading past the first paragraph of a ctrl+c/ctrl+v AI substack, it's a highly formulaic brain-rotting machine for the writers out there: "I do not want to add AI's patterns and tropes to my own thought patterns that come out in my writing, since reading is how people learn how to write."
Maybe readers should strike back at those "authors" who are dumping massive amounts of AI text under their byline by using AI summaries to get the gist of those articles. Then we'd have AI writing and AI reading, so the human is 100% left out of the picture! I am kidding, of course. I wouldn't use AI to summarize any more than I'd use it to write. I am an AI tee-totaller! But the scenario is cautionary. How much of our innate capacity to think (reading, writing and thinking are all aspects of the same cognitive function) do we want to lose by deciding not to use them?
I have been diligently striking back! I politely tell every author whose post is clearly written by AI that their post appears to be written by AI. Some take it in stride and confess. Some get huffy and dig their heels in. One author blocked me just for saying that! And she is apparently a truth and freedom loving person who doesnβt censor dissenting opinions on her substack - just embarrassing comments like mine
Good on you! I have said that to one or two authors using AI. One tried to tell me AI can produce wisdom! And this was someone who did not fall for any of the nonsense of the last 5 years and seems to see the danger of digital prison, CBDCs, etc. Yet appears to have fallen into the AI trap.
100%
Wow! Great piece in so many ways! Interesting analysis on Joshua. I subscribed to him liking the "topics" he was addressing. I read little of his work now as I find it "cumbersome". Cumbersome matching the various techniques the AI uses . I guess I demonstrate my naivete in wondering how some of these people could crank out the volume of work they did. I had assumed they, unlike myself, were simply dedicated career writers who did nothing but write. After reading your piece and looking at their "style" I think you pegged it. My background is history and my arena is filled with "court historians" and charlatans that spout agendas for the Crown's coin rather than pursuing the truth. As you have demonstrated, writing is equally contaminated. While I love the substack concept, you hit a major issue on the privacy and anonymity. While the "subscribe" digital currency model is "convenient" and guarantees substack their cut. it creates great vulnerability for the user. Most do not see this, as convenience dominates their thinking. Some of us have experienced the down side first hand, I do not know a solution for mass scale. I know I purchase books of writers I like and have done snail mail contributions to writers that have that option. I suspect most readers are too lazy for that option. To your request, I use Proton to include VPN and proxy server. I get all my articles to my inbox. You do great work and this was truly a needed masterpiece in challenging times.
It's going to be a challenge going forward. As much as the cynic in me says Substack will take the loot and not give a damn, there's a part of me that thinks humans who are not yet accustomed to such obnoxious formulas will soon detect them and get bored with them. Not just the accelerated upscale of 2x "essays" per day, but the machine droning shit writing quality. I dunno, maybe that's naive thinking and they'll prefer shallow pleasures in agreement with Alazons over authenticity and original ideas.
Thank you Celt, for the email info.
Much like the calculator decimated peopleβs ability to perform basic math, I fear relying on AI to generate content will spell the end of most peopleβs ability to write.
When I learned via Iain McGilchrist (https://iainmcgilchrist.substack.com/p/quantity-kills) the extent to which college students are using AI to write their papers (https://nymag.com/intelligencer/article/openai-chatgpt-ai-cheating-education-college-students-school.html), I realized this eventuality may be nearer than I imagined.
If the current trajectory remains unimpeded, weβre probably only a few generations away from the populace resembling the Eloi, which will make it all the easier for the philanthropathic Morlocks to harvest the herd.
I wonder if neurological studies have been conducted to determine if peopleβs hippocampus has been shrinking due to increasing reliance on smart phones and computers rather than our memories.
Le Bon, Boorstin, Ellul, Postmanβit sounds like youβve been raiding my βFavorite Books on Propaganda, Mass Persuasion, & Psychologyβ (https://margaretannaalice.substack.com/p/favorite-reads) :-) Meerloo would make a good addition to your canon.
βIn a world racing toward mass illiteracy, will humans ever know what Chicken Sands or Orebarger Rings are?β
π€£π€£π€£
βI only use Grammarlyβs free browser add-on for spell-check and nothing more and will contenueu to youtiliz it for yure benifet.β
πππ
Finally, I want to thank you for being honest and honorable about your own use of AI. I have never used it myself and donβt have any intention of doing so, but I must partly thank Emma Sron for warning me away from even trying it as a research tool (https://emmasron.substack.com/p/artificial-intelligence-ai-the-bigger). I donβt want to be complicit in feeding the beast data that may eventually be used to enslave us.
That said, I do feel a bit differently about AI multimedia. I still donβt intend to use it myself but have been deeply impressed with how effectively and rapidly Visceral Adventure (https://visceraladventure.substack.com) has been able to generate tyranny-detonating animations such as her rendition of my dystopian short story βIn Five Yearsβ (https://margaretannaalice.substack.com/p/in-five-years-video-blade-runner) and Noel Spanglerβs βJob Opening: Division & Chaos Lead, Global Elite Planningβ (https://visceraladventure.substack.com/p/the-satire-will-continue-until-the).
This is a case where I feel like we can use the tools of our enemies to help defeat our enemies as I discussed in a recent post (https://margaretannaalice.substack.com/p/50-reasons-to-give-your-child-video) in which I then quoted from my 2022 Corona Investigative Committee presentation notes, βA Mostly Peaceful Depopulationβ:
βThere is perhaps no more important tool for turning the tide of public opinion than framing, so if we want to win this war against the democidal dictators and their enablers, we must use framing, repetition, and viralization to propagate the truth.β (https://margaretannaalice.substack.com/p/a-mostly-peaceful-depopulation β’ https://margaretannaalice.substack.com/p/my-corona-investigative-committee)
P.S. Unbekoming has been so kind to me and appears to be a person of such impeccable character, I almost want to give him a pass for using AI to help with his exhaustive surveys, but I agree disclosure would be the right thing to do when authors do resort to such tools.
P.P.S. Regarding the tech shenanigans, I have been experiencing a ridiculous range of those myself, from people getting kicked off my mailing list without their awareness (https://substack.com/@margaretannaalice/note/c-92644006) to this recent scary warning Gmail slapped on my Billionairian Blueprint post (https://margaretannaalice.substack.com/p/billionairian-blueprint):
β’ https://substack.com/@margaretannaalice/note/c-146123251
For the record, Iβm on ProtonMail and did receive your newsletter.
P.P.S. I do wonder if your email is one that got knocked off my mailing list as I recently attempted to comp you but discovered you werenβt on there. If you want to resubscribe, Iβll be happy to apply a comp.
P.P.P.S. I apologize for the length of this comment, but you can blame your human-composed content for being thought-provoking and engagingβand catching me during a rare window when I wasnβt working on my own content as I was just preparing to crash after my third 20+-hour day in a row.
I am so grateful for this post! A true learning experience.
Having read several of Josh Styleman's postings I would not have had any idea he was using AI generated language. Even though I noticed he was for the most part telling what others (originals such Patrick Wood) have already described in a scholarly way about Technocracy, Styleman's writing flowed easily, and I read it quickly. I get emotional about what is happening to humanity--no longer being human?-- but fooling me is something I must learn to detect more proficiently. You have given me a gift. Thank you! I am not praising you for your admission of using AI in some way in some of your previous work, but it comes across to me like using other tools to get your ideas across in readable, comprehensible form. In the old days, a dictionary or thesaurus or rules on writing a good essay for example (I eventually advanced to spellcheck =:) ). Experimenting with AI to see what it can do is how you are able to teach us who seek knowledge about it. Thank you Professor! Lifting the words whether a phrase to a paragraph of others is abhorrent to me. AI will turn humanity completely against God's Creation as long as it is in the hands of those who's aim is to own the world.
If not using AI means I am an unsophisticated rube of the backwoods, that's fine with me. I'll talk to the trees.
Positively uplifting article, GC. Canβt wait to see all of this AI stuff play out. I want this to just be a stupid fad but only time will tell. Would you find it surprising that use of AI is encouraged by Navy leadership? It is considered a time saving tool.
Thanks, Jennifer. Doesn't surprise me. A perfect addition to other surveillance tools already in use. I suppose this has been the norm for over a decade. AI will probably send any sensitive content warnings directly to supervisors or up the chain. I will say for 'tech' and 'maths' and organizing workflow on projects, it's a game changer, but it all depends on the rewards for the user versus the privacy sacrifices of using it.
Good book, my friend, will have to get back to this one when I have more time!! This one will probably take a couple of Dos Equis to get through. Linking as usual @https://nothingnewunderthesun2016.com/
I use Hotmail and this did get through by the way.
I've known Unbecoming has been using Ai for a while. I ignore it because I look at them more like a summation service, and, itβs $4/month.
Yeah, I think that maybe the role they're happy to assume now with a higher paid subcount, but also with deteriorating quality. A lot of copy and paste dollar store substacks will be popping up across this platform in the next year, and it will be more difficult to find those unique voices.
You had me at the Killer's song "Human." I usually save myself a lot of time by getting bored with Substack articles before I can finish them which I realized in reading this were probably the AI generated ones.
I also save myself a lot of time by writing all of my own content based on thoughts currently circulating in my head. As I move into deeper spiritual based modes of inquiry lately I'm coming to realize that the only way to reach other humans is to be human...
God Bless you for this, and May He Bless us all with Clarity.
This follows our world-wide soul-crippling weaking and weakness as outline in my article here. Worth your consideration.
"Multiverse Journal - Index Number 2225:, 21st July 2025, The Profound Sickness: An Anti-Abortion Apologetic"
https://stevenwork.substack.com/p/multiverse-journal-index-number-2225
Feed-back welcomed.
God Bless., Steve
Some 'Salt' into this exposed wound?
Listen to the AI generated overview 'conversation' of this - your - article. {Right-side, lowest or next lowest, ~45 minutes}
-
Google NotebookLM;
https://notebooklm.google.com/notebook/ad2d217c-48d4-4d78-838e-46a580c11ca9
Recently, I asked Perplexity what I would need to take with me for a trip to Neptune. It gave me a list of things to prepare and have along as if this was an everyday occurrence. I guess it assumed that man has gone to Neptune for a visit. Is the government hiding more crap from us or is A/i as retarded as it can be?
Answer is C. Government is hiding from us just how retarded AI really is. When this bubble pops, it will make pets.com look like a solid investment in 1999.
The pop will likely resemble the words of the illustrious Condaleezza Rice: βWe do not want the evidence of Saddam Husseinβs weaponsβofβmassβdestruction to come in the form of a mushroom cloud.β The mushroom cloud will be the meltdown of nuke plants that are being built to supply power to AI machines. My own sense of things is that that is the point of AI, build something that requires lots of power and use that as the excuse to build nuke plants, then melt them down and depopulate the area except for wildlife ala Chernobyl. I live 5 miles from three mile island which was shut down, I foolishly thought for good until they now are going to get it going again to feed AI and data farms. Data farm being built on the site of a really nice municipally owned golf course. So, as you point out, when the bubble pops, it will really pop. I have to admit I had to use the proton AI "Lumo" to remember Rice's name because I forget everything now, and to get the quote right. Oddly enough when I asked it for her quote about mushroom clouds it gave me this: βWe do not want to see another mushroom cloud over any city, anywhere in the world.β, this quote was from an obscure speech she gave to a bunch of generals and spy types. I had seen her repeat that quote so many times I knew that wasn't it so I had to paraphrase it and then it gave me the right quote. So, did it try to sanitize her infamous quote with one a little less innocuous? I'm very disappointed now because I cannot discern the AI writing you refer to in your article so I will be dependent on yet another stupid machine to figure it out.
This article is one of the reasons I became a paid subscriber, you hit on a lot of things about this world that concern me. Another brilliant piece of writing, who needs AI when people like you still exist. Thankfully I normally realize when a real person is behind the writing and tend to know when someone is no longer producing their own work but has resorted to AI instead. Can't stand it personally, I don't want speed I want writers who have to think about what they are saying and I do want depth.
The rot for me set in long before people had access to computer's and mobile phone's. Back in 1987/88 all parents where I was living at the time received a letter from the schools notifying them they had to provide their children with a calculator for their maths lessons, not a cheap item at the time. No more having to prove the workings of the maths, just tap it out on the calculator, job done. Believe it or not I have only ever used a calculator once in my life and because I couldn't quite trust it I still ended up working it out on paper to check it was right. It was but I realized it just wasn't for me.
By the way, I absolutely loved your " Shameless human-created cyberbegging interludes".
Wow, you and I are among the 12 subscribers left!
One for each month!
Assuming you are being serious that's terrible, our Good Citizen is the one that stands out, his ability to write intelligently, seriously with that thread of humour running through his writing's and not afraid to say things exactly as they are is worth a gold mine.
Admittedly I had to save to get a yearly subscription but that was more because I dare not do it monthly as here in England the DWP have had access to look through our bank accounts for about a year now to see what we're spending our money on and our government has slowly started chipping away at our retirement pensions in various ways. A monthly payment may have red flagged it as a needless outgoing.
I guess I could have gone for pensioner option but he's worth full payment to me, I'm already saving to upgrade .
Using machines to monetize the consoomer demand for more Content(tm) criticizing the use of machines strikes me as less likely a grasping toward authority and more a reddit smuggie move.
Stupidity is well-distributed across worldviews and ideologies, and the stupid often self-identify by claiming their camp has only smart people (though to be fair some have few or none). So, far from a clever move, monetizing the idiots is dimwitted and self-degrading.
I can think of few admissions-by-action more embarrassing than letting an LLM "improve" one's writing. Maybe a link to an onlyfans account.
Speaking of, don't let LLMs do math for you. They're worse at that than most things they're terrible at. I'll never stop laughing at LLM devs struggling in vain to get massive piles of linear algebra to do basic arithmetic at a millions of times the computational cost of bare metal.
If you need some help with obnoxious math, try WolframAlpha. Its syntax is less tolerant than an LLM but it's a graphing calculator on steroids that runs on almost-plain language and, crucially, actually does the math instead of guessing the answer.
Hahaa. I always need help math. The last course I took was in high school, and it was my only C over four years. I used the upgraded versions of GPT (4 maybe) for statistics calculations on an EV post nearly two years comparing mpg of a Tesla and Honda Accord hybrid ago and didn't check its work. Damn, I hope it wasn't a disaster.
You might also try using AI-assist in spreadsheets. I canβt vouch for their capability (havenβt tried), but at least if the LLM generated a bad formula the calculations would be consistent haha.
I pray that this is written by a human being. Long live the writer and storyteller.
Epic Piece, GC. Thank you for all of those carefully-explained AI markers.
We have no doubts about your integrity. Unlike most folks in these times, you know that no one can give to a person his or her Self Respect. People need to 'Take It.....For Themselves.' And, people without self-respect are incapable of respecting anything.
Many decades ago when I was a young lad at Old MacDonald's Farm located at Harmony Junction on Prince Edward Island, local persons who told lies were publicly shunned. Such treatment was highly effective in a time & place with no electricity or home telephones and the nearest "Mountie" was an hour's drive if you had a motor vehicle.
If Samuel Clements were alive in today's JUSA he might write: "Can't swing a dead cat without hitting a paedophile."
"How Green Was My Valley."
P.S., Your Msg. was Rec'd via Hotmail/Outlook and, Bill Gates likes neither of us.