AI and the knife collector.

dsutton24

Gold Member
Joined
Apr 9, 2018
Messages
2,045
We've all seen them.

"This xyzt knife is a reliable and stylish addition to any collection. The stainless steel blade is durable and the manual opening mechanism provides easy access when you need it. The knife's brown color and plain edge give it a classic look that is perfect for any outdoor adventure..."

They stick out like a sore thumb. They tell you nothing about the knife whatsoever, but other than their style they don't raise any of the old alarm bells. Chances are the seller isn't a knife guy. It might be one of those folks who troll through flea markets and thrift store looking for things to resell at a profit. It might be whatever a Boiler Room operation is called these days, dozens of people working to unload a container full of knockoff goods.

In recent years you have seen more and more "pic is description" merchandise descriptions. I always look at those as basically a crap shoot. If the photos are good enough I just rely on what I've learned about the subject and decide whether or not to take a chance. It's hard enough not being able to handle an item. Figuring out what you're buying from online photos is an art in itself.

<warning: soapbox>
And the counterfeits are out there. The number one tip you see is something like, "Why would anyone counterfeit a $20 knife when he could (whatever)?" just doesn't work any more. Please! Stop spreading this one around. It makes perfect sense, but it's at least twenty years out of date. They counterfeit toothpaste and Sharpie pens, for heaven's sake. If you don't believe this, go to the auction site and search "Buck Portable 110'.
<end: soapbox>

Anyway...

AI is a fact of life, and right now you can spot it a mile away. It's fun, it's funny. But, look at how much it has improved just over the past year or so. AI will continue to improve. The writing style will evolve beyond the vague, cookie cutter prose to something that is more conversational. It will also become compelling and persuasive. And the information will have an agenda. You just bought a thousand knockoff Case knives? You create an AI project that will appear to give them provenance going back decades. A hint on one forum then another. An apparently old reference to a catalog from the past. A mix of articulate, authoritative information, casual chatter, and even a little controversial or spurious content. Bam! Instant credibility.

The majority of the folks living today have had the Internet in their lives. A lot of these folks have an alarming lack of skepticism. There was a thread recently where someone came along looking for some reference materials on Traditional knife patterns. I poked around briefly on the web and didn't find anything I liked, and then just fell back on Levine's and the old materials that are long out of print. I'm not going to wade through a thirty minute video when I can find what I need in thirty seconds of reading. But then again I'm very old and can be a pain in the neck.

I guess this is the point. How do we make sure that good information is out there for someone who is interested? If your concentration is on whatever you can buy today, then no problem. There's a ton of information out there of a quality commensurate with your twelve dollar knife. Suppose, though, someone wants to make a logical step. How does my knife compare to something that was carried by my Grandpa? What choices did he have when he walked into a hardware or sporting goods store? Why would anyone pay big money and search for years for the same knife that can be had today for twenty bucks?

As wonderful as this place is it has its limits. It certainly figures largely in search results.

Is there a way to make sure good information that is relevant to the question asked is out there?
 
This really will become ever more of a problem on the internet. If you look around you can find articles about the "ens**tification of the internet" (I think you can figure out the asterisks). The ens**tification of the internet is going to make it very difficult to figure out what is real and what is just AI/algorithmically generated garbage. Unless the big tech companies decide to somehow curb their AI creations, I fear that we may have already seen the best days of the internet.
 
There is a fascinating theory called the “dead internet” that asserts within a few years 90+% of all internet posts and traffic will be bots and AI talking to each other, and we are very close to this already.

 
There is a fascinating theory called the “dead internet” that asserts within a few years 90+% of all internet posts and traffic will be bots and AI talking to each other, and we are very close to this already.


There was a Pew Trust study that said, basically, that about a quarter of the internet disappeared over the course of ten years or so. Things like local papers, blogs, and things like that just vanish. It takes money, time, and more money to get this stuff on the web, and then somebody has to keep it available... more money and time.

I have absolutely nothing to back this up, but it seems likely that useful stuff disappears more quickly than the fluff. I've got a few NKCA annual knives that I've bought over the years, I never really got around to joining. At one time the NKCA was huge, but I found that it was very difficult to find information. About the only thing you find are sale listings. If the NKCA can basically vanish without a trace how will somebody a few years from now learn about the arcane stuff that we chase?

There's a series of ads for an insurance company that makes fun of people turning into their parents. "You don't have to print the internet," he tells us. Don't believe it. If I find a useful bit of information it gets filed away on a hard drive that gets backed up regularly. I get as much out of finding history and folklore as I do in just handling knives.
 
AI is a fact of life, and right now you can spot it a mile away. It's fun, it's funny. But, look at how much it has improved just over the past year or so. AI will continue to improve. The writing style will evolve beyond the vague, cookie cutter prose to something that is more conversational. It will also become compelling and persuasive. And the information will have an agenda. You just bought a thousand knockoff Case knives? You create an AI project that will appear to give them provenance going back decades. A hint on one forum then another. An apparently old reference to a catalog from the past. A mix of articulate, authoritative information, casual chatter, and even a little controversial or spurious content. Bam! Instant credibility.
I have been using DuckDuckGo as a search engine for years, and I noticed recently that if I search for information on some random topic, almost all the top hits are pretty clearly AI generated garbage. The websites all have ridiculous names like YourPlumbingAndLawnMaintenanceHelpInformer (ok, I made that up, but you get the idea).

If AI “learns” by crawling the internet for information, and more and more of internet content is already AI-generated (with all its inaccuracies and stilted language), I wonder how long it will be before all internet content is unusable gibberish. I’m probably just displaying my ignorance of how these things work, but still, I wonder…

I just now searched for something I thought likely to generate AI hits: “Best knife handle materials”, and pretty much got what I expected.
 
There was a Pew Trust study that said, basically, that about a quarter of the internet disappeared over the course of ten years or so. Things like local papers, blogs, and things like that just vanish. It takes money, time, and more money to get this stuff on the web, and then somebody has to keep it available... more money and time.

I have absolutely nothing to back this up, but it seems likely that useful stuff disappears more quickly than the fluff. I've got a few NKCA annual knives that I've bought over the years, I never really got around to joining. At one time the NKCA was huge, but I found that it was very difficult to find information. About the only thing you find are sale listings. If the NKCA can basically vanish without a trace how will somebody a few years from now learn about the arcane stuff that we chase?

There's a series of ads for an insurance company that makes fun of people turning into their parents. "You don't have to print the internet," he tells us. Don't believe it. If I find a useful bit of information it gets filed away on a hard drive that gets backed up regularly. I get as much out of finding history and folklore as I do in just handling knives.
Yup, I am coming to this conclusion too. I went to find info on a site about Robeson knives the other day, and the old website that I had used before with a treasure trove of info is no more. So yeah saving a PDF copy of websites does really seem to be the ticket nowadays.
 
Yup, I am coming to this conclusion too. I went to find info on a site about Robeson knives the other day, and the old website that I had used before with a treasure trove of info is no more. So yeah saving a PDF copy of websites does really seem to be the ticket nowadays.
I could not find anything on my latest Western L66 search to re-read some of it, either.
All my Robeson searches (for example: " Production years of the Robeson 622110" with and without "english jack"; "large jack knife"; and "folding hunter" brought up nothing but irrelevant results.

Ditto for "Slipjoint (and "slip joint") folding hunter clasp knife history" Mainly results pertaining to the Buck 110. 🤨👎

I suspect information on pre-Queen Robeson models and production years was slim to begin with, and has only gotten (much) worse. 🤔☹️
 
Last edited:
I could not find anything on my latest Western L66 search to re-read some of it, either.
All my Robeson searches (for example: " Production years of the Robeson 622110" with and without "english jack"; "large jack knife"; and "folding hunter" brought up nothing but irrelevant results.

Ditto for "Slipjoint (and "slip joint" folding hunter clasp knife history" Mainly results pertaining to the Buck 110. 🤨👎

I suspect] information on pre-Queen Robeson models and production years was slim to begin with, and has only gotten (much) worse. 🤔☹️
I believe you are right. I tried finding pattern numbers for Robeson for ages and the only part I could find was from the old Robeson site that was pasted into a post on Bladeforums here. Sad stuff indeed.
 
Back
Top