[jvandonsel]

He seems to be doing everything by "feel". I would imagine that if the stone were 3D scanned it wouldn't be difficult to use some software to determine the optimal orientation to achieve the largest cut stone, for a given cut style.

[jasonwatkinspdx]

He's making decisions based on a lifetime of experience. I wouldn't be so quick to assume that can be replaced by a simple optimizer. Additionally, maximizing size is just one goal among many.

[dylan604]

The hubris of any software dev is that code can do anything better. For a large majority of things, I would agree. After all, that's why we made computers in the first place. It's also my tenet of workflow automation. I'm not trying to eliminate a person's job. I'm just trying to make the tedious error prone parts of the job less tedious and error prone so that the person is free to spend more time on the things that they are good at and would be difficult to automate.

[jasonwatkinspdx]

Automated gem cutting exists, but is only used for lower value stones. If it were such an obvious slam dunk the industry would be using it for stones like this too.

I like to use Toyota as an example, because they famously reject total automation as a desirable goal. The believe having humans in the loop creates a learning loop that gets the best overall result, so they optimize for a sweet spot between automation and human.

To use an analogy from art, automation doesn't replace the artist, it replaces more limited paint brushes. As amazing as recent large model advances are, there's also a clear difference in how effective humans are at using them to get the desired result.

If you're just playing around then a simple prompt to finished image pipeline is convenient and shockingly good. If you're being more specific about the end goal then you need the model to expose more fine grained ways for the user to express what they want and modify the intermediate results. This why you see Adobe et all focusing on that kind of "smarter tool used in detail" approach.

[NikolaNovak]

>>If it were such an obvious slam dunk the industry would be using it for stones like this too.

I don't think so, or at least I think it misses where this industry drives value from. I don't know if my comment will seem too cynical for the Hacker News add value to discussion threshold, but - the entire industry is by definition built, completely, around sub-optimization and artificial complexity, scarcity, and meaning. If you want optimal and efficient and automated, you can get it for fraction of the price. The hand made sub-optimal facet is a strong part of the perceived / advertised value.

[jasonwatkinspdx]

Yes, because simultaneously it's somehow true that it's a slam dunk to use automation to replace this skill, but yet industry faces some sort of irrational barriers to implement that.

Or maybe, just possibly, armchair opinions from programmers who have never done this and have no idea what the reality is beyond a youtube video have superficial understanding?

Which do you think is more likely?

[tgsovlerkhgsel]

A cheap quartz movement (watch mechanism) will be more accurate than most high-end mechanical watches, and yet they keep being made.

I think the main reasons (beside "vanity") are:

a) huge profit margins, limiting the incentive to optimize

b) high start up costs combined with relatively little volume

A system would have to accurately measure imperfections/inclusions, then pick the best design, then it'd either have to cut it themselves or communicate the instructions to the person (who'd have to be willing to use such a system).

The current approach with a human designing and cutting the stone is simply good enough, so there is no reason to change. And something like re-cutting the Koh-i-noor happens rarely enough that I suspect nobody wants to invest in developing the software for it when you can get a close enough result by throwing more one-off manual work at the problem.

Synthetic stones (which I assume are a much simpler problem because they're even cheaper and have fewer inclusions to optimize around) are already being cut by robots (or so I've been told, at least). I think the whole "real natural stone" and "hand-cut" parts are definitely part of the appeal/selling point.

If you care about a "slam dunk", you'll buy robot-cut Cubic Zirconia or Moissanite and call it a day.

[NikolaNovak]

I have absolutely not posited the barriers as "irrational". Artificial market segmentation, and increasing value of product through human artistry are increasingly common and not necessarily "irrational". There are any number of products on the market that machine can make better, but we value more highly when it's made by human... in fact, that's true of a very large proportion of products, marketed as "handcrafted" or, since it's 2024, "artisanal and rustic" et cetera :-)

(the line of where "handcrafted" goes from "friendly and community and positive and encouraging neighbourhood entrepreneurs" to "evil and scheming and big-corpo-deceptive" is subjective evaluation, but none of them necessitate "irrational" let alone, as is implied, "stupid")

[dylan604]

Anytime something takes a long time to create, there's an expected add to the price. Look at your 18 year old (and older) Scotch/Whiskey prices. Even lab grown gems takes months to "grow". Yes, there is a cartel influence on an artificial scarcity, but even still, it's something that isn't just common everyday discoveries for the vast majority of people. Even without the cartel influence, there would still be value in them vs much more common things like quartz/granite/etc.

[nickburns]

'computers' can now rival the greatest painted masterpieces from all of human history with a mere text-based prompt (or less). would you consider those generated works to be masterpieces? or is there something about a work created by human hand and derived from a human mind that can only make them so?

[abeisgreat]

The thing with masterpieces is that they don’t hold that status because they can’t be replicated, they hold it because they were novel, innovative, and unrivaled at the time of their creation.

[shagie]

> The thing with masterpieces is that they don’t hold that status because they can’t be replicated

Copying the old masters is often an important part of developing one's own skills as a painter.

https://www.sightsize.com/old-masters-copying-older-masters-...

> The training of painters in past centuries regularly involved copying old master drawings and paintings. To that end, most museums happily allowed students into their buildings for that purpose (some still do). But the practice was not limited to students. Even fully trained old masters copied older masters.

Why copying old master paintings is useful - https://youtu.be/91UXW_hSpnU

The Art of the Copyist - https://www.metmuseum.org/perspectives/videos/2023/3/copyist...

Art: France’s long history of copying Old Masters at the Louvre - https://www.connexionfrance.com/article/Mag/Culture/Art-Fran...

> The practice of copying and recreating paintings by the Old Masters at the Louvre goes back to when the museum first opened in 1793, when any artist could turn up and use a freely available easel to copy a masterpiece.

> ...

> Not all artists copied works to improve their skills. Some took up the practice professionally, since the demand for copies of masterpieces in the Louvre was high throughout the nineteenth century.

> ...

> These days only 250 copyists are permitted to install themselves in front of the museum’s art works, and a two-year waiting list shows that there are plenty of hopefuls waiting in the wings to take up a palette and brush.

> Those granted access have up to three months to work on their copy.

[jasonwatkinspdx]

That this exists as a pedagogical exercise does not disprove the original point in any way.

Source: I spent a lot of time in the library copying sketches of the renaissance masters as a kid.

AI is the pencil, not the artist. As cool and capable as large models are they are not even remotely close to replacing self directed human intent. If you do not understand this you do not understand art.

I don't believe there's some magical quality to human intelligence, just that the things we are making today with AI are still orders of magnitude short of the real thing, and that there are still very difficult open questions in that gap.

[h0l0cube]

Almost every time I see an AI generated art piece, I know it's AI generated, and it looks cheap. Sometimes an actual experienced artist uses AI and manages to tweak it to look decent, either by prompt engineering or further manipulation. There's nothing that yet replaces the human taste for what looks good nor what resonates for contemporary audiences. At best, it could be said that AI has further reduced the value of derivative art and photography.

[leptons]

A lot of the AI "art" I've seen looks like a psychotic serial killer made it. There's just something wrong with it, especially when it depicts humans - the eyes are insane, the facial expressions look like the subject wants to cannibalize me. If it were a human artist, I would suggest they seek mental help. It's honestly kind of scary that this is what computers are creating, it's no wonder some people worry about AI destroying humanity.

[6510]

One time I only input the style of the image basically leaving the description empty. I was wondering what it would come up with by it self. It served a human head on a table. I didn't need a second image.

[bingbingbing777]

Hardly. They cannot mimic the brush strokes, the material choices, expressiveness, composition and more. It is absolutely abhorrent to say that any AI is even in the same field, let alone close to art masterpieces.

[leptons]

AI "artists" don't even seem to know how many fingers humans are supposed to have.

[TaylorAlexander]

> 'computers' can now rival the greatest painted masterpieces from all of human history

This is only true if you misunderstand the meaning of art as a concept and its relationship to humanity.

[nickburns]

my provocative phrasing was of course intentional (which i'm all but certain you realized). couldn't agree with you more.

[TaylorAlexander]

This was not at all clear to me, nor is it clear upon rereading it now with this information. Plenty of people believe that without reservation.

[wizzwizz4]

Photocopiers can also rival the greatest painted masterpieces from all of human history.

[nickburns]

nah, they just plagiarize. (for which, coincidentally, the same could be argued for human-made and trained AI.)

[fullspectrumdev]

> is there something about a work created by human hand and derived from a human mind that can make them so?

Yes.

[nickburns]

my downvoted-to-purgatory comment you've quoted. glad to see that the phrasing of that specific question stuck out to at least someone in the manner i intended.

the answer truly is a simple 'yes.'

[WXLCKNO]

You can say the same thing about chess/go. A lifetime of experience didn't help people getting beat by computers.

[mitthrowaway2]

My guess is a skilled craftsman could get within 3% of the optimum solution.

[MichaelZuo]

I would agree for all the common shapes, maybe not for some very odd asymmetrical shape.

[n1b0m]

A video posted by another commenter shows someone using a 3D scan of a gemstone to work out how to get the most out of the gemstone he is cutting

https://www.youtube.com/@hedronrockworks/videos

[jccalhoun]

This is done for some diamonds. About 1:20 in this video you can see them mention how they use computers to aid in deciding how to cut the diamond https://youtu.be/8lk8p0re8Eg

[nickburns]

great point... but in thinking about your comment, it occurs to me that then it could no longer be considered "hand cut."

i found it very interesting that he shared he was hoping for 300+ cts., but had resolved to be satisfied with at least 250. so he did alright after all.

[rkagerer]

294.5 cts in the end.

[iancmceachern]

They do this for larger diamonds, it wouldn't make as good of a video though