Plate AI — Plating The 1851–1857 3c With Artificial Intelligence

For decades the cost in man-hours of plating the 3 cent imperforate far exceeded their market value. It was purely a labor of love. PlateAI will greatly improve that effort/value gap and the fun factor will increase.

Can I plate this ugly unloved stamp? Seeing if you can actually plate those unplatable stamps could become a thing. Those of us holding hoards of unplated stamps that are "maybe plating quality, but I will get to them some day" will now have fun plating them and put them into the marketplace.

One of the hesitations to attempting to plate a pane of the 3 cent stamp is the probability of filling your gaps. I believe that will change as much more sidelined product becomes available and you can confidently learn to plate your own stamps. Almost anyone should be able to affordably complete a pane - solve that jigsaw puzzle.

I created the beginnings of a bare bones bourse/exchange capacity at https://stampplating.com/etcetera to allow 3 cent imperforate enthusiasts to have an "off-market marketplace". There has been zero action there. With PlateAi I can see this turning into something useful; and, if interest develops I can readily expand its features and benefits. Your ideas can help me do that. I will never commercialize it - it will always remain just a way for those who have this stamp connect with those who want this stamp.

I plate and collect 1c Franklin
I have found immense enjoyment in reading Neinken and studying on how to plate
I hope AI helps the hobby, and does not hurt it

I went back to philately due to the challenges of it, history, and cerebral nature of examining a specimen

I left sports cards because it was boring

Hopefully this AI stuff makes hobby better… but there are downsides

I was somewhat worried about how 3¢ collectors would respond to having a robot that can do what they've painstakingly done for decades. . .I am optimistic that it will increase the enjoyment of the hobby - the handful of beta testers are serious students of the 3¢ and proficient/expert platers. . .so far the feedback I've heard has been overwhelmingly positive from that small sample size. I'm also optimistic that it will increase interest in this issue and in stamp collecting generally.

For this tool, specifically, here's an example: I have several Type I, A relief, no recuts, average-looking frameline plate 6 or 7 3¢ Washingtons. I have spent hours trying to plate them, but that work is super boring and I cannot focus too long on boring stuff before moving on to something else. It was frustrating because I know those stamps were platable - nice copies, 4 framelines - not stockbook rejects. It hurts my OCD tendencies to put a 3¢ stamp in my main collection that says "unplated." With this tool, 4-margin stamps can be plated in minutes (or less) - if not the model's top choice, it is probably in the top 5. I still have to check, but instead of looking at hundreds of possibilities, I'm looking at 5 - and instead of playing with multiple images - lining them up, maybe making a compression scan, organizing them on my screen to compare to references. . .I spend all my time on direct comparisons.

I am a relatively young collector (under 50), and I think the real embrace of technology is a key to engaging new collectors. Tools like stampplating.com for the 3¢, Richard Doporto's 1¢ archive, Siegel's power search, SAN, StampSmarter, Jim Forte's post office lookup, PSE's SMQ and POP, Stamp Community, and other similar tools and sites make stamp collecting a better experience for me - not to mention the LLMs that I've found to be an amazing (though not totally trustworthy) companion to (reading and) understanding covers. Most, but not all, of these tools were created by collectors without a profit motivation, and with barriers to using advanced technology dropping at accelerating rates, I expect and hope to see more. I think it is inevitable.

If all goes well, I'll release it on Sun or Mon. . .

"Given enough training data, plating, stamp identification, and some kinds of stamp fraud"

As long as it is fed by experts and doesn't scrape the internet as there is plentyof what you scrape of the bottom of your shoe out there.

edited to add: In this case I am comfortable with the feeding of this particular AI project.

Well done gentlemen. Thanks to all who worked on this. I've only used it briefly so far but it looks very promising. One downside is that I suspect fewer 10/10A in the lots of 11/11A I buy online. #128579;
Regardless, its a very nice tool and im grateful you made it available for folks to use for FREE.

Regards,
Harper1249

I have some interest in the 3 cent stamps and went to the plating tool and found plating AI live. I gave it a try and am very favorably impressed. I am pleased to see this discussion here which ratifies that it is not a scam and has real credentials as to its' origin. Having recently obtained some new to me 11's this will be most useful.