AI art knows Spyderco

[Naperville]

I have used GPT-4, Bard/Gemini, and Copilot for many hundreds of hours, not to design anything but to answer questions about healthcare, heart and diabetes medications, supplements that might help with heart problems, diabetes, and in the performance of living to 100.

Right now, as far as text based queries go, AI is nothing but a large language model(LLM). They "train" the AI by feeding it documents. You ask it questions....THAT IT DOES NOT UNDERSTAND, and the LLM predicts based on the query what the words are to respond.

It is a prediction engine. That is the my understanding of AI at the moment. How the graphical AI systems work I do not know, maybe it is the very same.

At the moment, today, NOW, any AI that is open to the general public is really just an enhanced search engine, with ONE lil difference. They TRAIN it by feeding it a particularly chosen set of documents creating BIAS, and make "TAGS" or queues for words to follow certain words or ideas based on your query.

He who TRAINS the AI is in control of the ideas that the AI system puts out. It is an idiot machine at the moment. It saves me time in searching for the documents and reading them.

I do not think it is a great idea to take the shortcut of a summary document and haven't used it as extensively as I had, because I like to see the documents and do my own research.

AI is not really producing ANYTHING at all that is new. It does not have that capability yet. It does not think, it regurgitates what it was trained on. If the people training the machine are all Marxists, then all you are going to get are Karl Marx type responses.

[shunsui]

Art as two dimensional drawing/painting is often theorized to have died back around the turn of the 20th century. Lots of arguments. One that comes to mind is photography killing the need. Another is the whole intellectual (Marxist ?) boredom with a precious thing you put a frame around and hang on the wall. What remains is simply commercial illustration.

Art or illustration as ALL trades are subject to the economy. Time is money. If AI can illustrate something in 0.037 seconds, and someone costing the company payroll $60/hr needs a day to do the same work, someone is going extinct.

When I joined a manufacturing company back in the 80's, we had a lot of guys doing drafting with paper and pencils. Five years later they all used AutoCad. Ten years later it was all Computer Aided Design / Computer Aided Manufacturing (CAD CAM).

My recommendation for anyone that feels their Artistic career is threatened by AI is to learn how to use it or find a new career.

[RustyIron]

I think you have me in the wrong box.

Sorry if I put you in a box, Sal. That wasn't my intention. On virtually any issue, I like to examine it from all sides, like a new Delica owner examines his knife the first time he takes it out of the box. Only through close scrutiny can we arrive at the truth, a better understanding of our competitors, and our own weaknesses.

[sal]

No Problem RustyIron,

Just wanted you to know that, as an old guy, I don't fear actual progress. Right now. A.I. is a wild horse. Some parameters will grow in the future.

For discussion, 45 years ago I designed and produced a folding knife that had a hole in the blade so you could open it with one hand and we put a clip on it so it could hang off the rim of your pocket. two years later, I designed a knife that had serrations. None of those features were being used at the time. It was not an easy thing to do. Now those features are commonplace in the pocketknife industry. Most knife buyers don't know that Spyderco brought them those concepts. I'm concerned that A.I. will shorten that time considerably, if not harnessed. This could dampen the "Spark of creativity". In my humble opinion.

sal

[Mat_ski]

The fundamental question people should be asking is:
Is it net positive or net negative?

Just be honest with yourself when answering this.
I hardly see any positive.

[kriezek]

It sure does seem like there is a lot of fear and hatred of new technology among Spyderco enthusiasts.

Automobiles,
Aircraft,
Computers,
Printing presses,
rifled barrels,
mirrors,
forks,
flushing toilets,
sewing machines,
the wheel,
aqueducts,
surgery.

Upon their introductions, the common folk were all terrified of these newfangled instruments of the devil. They protested with their pitchforks and flaming torches in the middle of the night, yet the brilliant humans prevailed, and the world is a better place.

Personally, I've used these AI computer interfaces to my advantage. Have any of the Spyderco users who are complaining so vehemently been injured or had their feelings hurt by ChatGPT, or is all the panic unfounded?

In computer parlance, GIGO (garbage in, garbage out)... AI has lots of potential to create VERY skewed views or SLIGHTLY skewed views that can change society as a whole. And that is where I have an issue. Just look at the Google Gemini fiasco where it was so blatant and obvious that they had black Nazis and Norse vikings, but no white ones in their photos. It is the answers that are more difficult to determine if they are NOT telling or showing the truth that is of grave concern.

The other problem is the reliance on AI when making decisions that should be made HUMANS. AI does not take into account the human factor. And that is the most dangerous issue. If AI is put in control of systems that can cause massive destruction or disaster to humans, then it is very likely at some point, AI will cause massive destruction to humans. Think control of weapon systems, or nuclear power plants, or even making decisions about transportation that doesn't take the human factor into account. I'm sure there are lots of other things I haven't even thought of.

Am I concerned about AI in manufacturing processes or designs? No. Am I concerned about using it in a limited capacity? No. But knowing human nature and how stupid and vapid people in power are, it will take a huge event to occur before people wake up and by then it may be too late.

Just because something can be done does not mean something should be done. But this is a knife forum. And AI with knives is likely involved in design and manufacturing. And I foresee issues with copyrights, but also some potentially really strange knife designs that may catch on with consumers.

[Scandi Grind]

I see AI as a tool, and just like any tool it's use or misuse will be determined by the user. It is just like a knife, it can be used for all good things, or it can be used to harm. The problem I see with AI is not unsimilar to many of the problems I see with the introduction of the internet. It is very powerful technology that relatively few people ended up learning to use responsibly. That is a broad statement, but I believe the internet is misused more than anything else, which is not unlike most tools humans have invented. But it really comes down to people, not the tools. You need good people to use powerful tools for good, but right now the people who are exercising free reign of the technology are using it primarily for corrupt reasons. To make matters worse, most will follow along with whatever such corrupt individuals cook up and end up being a part of the stew.

It can't be restated enough though, we actually need good people to wield these new tools powerful tools. This is no different than me believing in teaching kids the importance of using any tool, a hammer, a knife, a gun, or the internet, it is important that they learn to use these tools safely, morally, and yes effectively, even if that tool is AI. But before that, they must learn to be good. It always comes down to people, I'm getting philosophical perhaps, but I doubt any will disagree that nothing good will come from AI if there are no good people to wield it, or if we allow only the corrupt to wield it. Hearts and minds my older brother always says, every problem we have starts with the hearts and minds of people. That is where the battle lies, even though our technology changes the terrain that we fight on, the people and not our tools will determine what we will suffer or gain from the tools we are now being introduced to in this wild technologically advanced period in history.

I believe we should learn to understand AI and use it responsibly, use it for good. If we don't use it for good, it will only be used for bad, and it will be used against us either way. Don't bring a knife to an AI fight. Wield it, and wield it well.

[zhyla]

The fundamental question people should be asking is:
Is it net positive or net negative?

You pose that question like it affects the outcome. These software tools aren’t going anywhere.

[Naperville]

The fundamental question people should be asking is:
Is it net positive or net negative?

Just be honest with yourself when answering this.
I hardly see any positive.

It's a new fangled contraption so everyone wants to try it. As I said, I tried it and it is an above average summary device for topics you may be too busy to research yourself and don't care about the accuracy of the summary.

I'm far more worried about where we are headed with quantum computing destroying crypto, systems hacking and where AI is going say 5 to 10 years down the road. In a decade all of these technologies are going to upend everything and could cause wars.

Don't say that I did not warn you.

[shunsui]

...
When will the first AI get a patent?

sal

https://www.theverge.com/2024/2/13/2407 ... e-guidance

"In 2020, the US Patent and Trademark Office (USPTO) ruled that only “natural humans” can apply for patents after it denied a petition from researcher Stephen Thaler. Thaler added the AI system he created, DABUS, as an inventor in a patent application. A US court upheld the patent office’s decision. A different federal court ruled that AI systems cannot be granted copyright, following a separate application by Thaler involving an AI-generated image."

There's more though.

https://www.ropesgray.com/en/insights/a ... pto-speaks

"While AI systems and other non-natural persons may not be listed as inventors on U.S. patents and patent applications, the use of an AI system by a natural person will not preclude that natural person from qualifying as an inventor (or joint inventors) if the natural person significantly contributed to the claimed invention (as discussed below)."

[Mat_ski]

The fundamental question people should be asking is:
Is it net positive or net negative?

Just be honest with yourself when answering this.
I hardly see any positive.

It's a new fangled contraption so everyone wants to try it. As I said, I tried it and it is an above average summary device for topics you may be too busy to research yourself and don't care about the accuracy of the summary.

I'm far more worried about where we are headed with quantum computing destroying crypto, systems hacking and where AI is going say 5 to 10 years down the road. In a decade all of these technologies are going to upend everything and could cause wars.

Don't say that I did not warn you.

The “AI” are already capable of plenty more. It is hot topic in civil engineering industry. There are models being “trained” to make CAD drawings. How do you feel about structures or bridges “designed” by an “AI?”

[zhyla]

How do you feel about structures or bridges “designed” by an “AI?”

Ambivalent. Are you concerned the software will make more mistakes than humans? Humans make a lot of mistakes.

Such a strange turn this thread took. Is everyone in some kind of mad panic because of some software hat tricks all of a sudden? Software has been swallowing industries for decades.

[Mat_ski]

How do you feel about structures or bridges “designed” by an “AI?”

Ambivalent. Are you concerned the software will make more mistakes than humans? Humans make a lot of mistakes.

Such a strange turn this thread took. Is everyone in some kind of mad panic because of some software hat tricks all of a sudden? Software has been swallowing industries for decades.

Yes, I am concerned software will make mistakes, but the worst part of it is that the responsibility for mistakes is removed.

[zhyla]

Yes, I am concerned software will make mistakes,

Are you concerned humans will make mistakes?

The concern is understandable but keep in mind software is infinitely more testable and verifiable than humans are.

but the worst part of it is that the responsibility for mistakes is removed.

I’m not sure how to interpret this. The responsibility does not change. The company that builds the bridge is still responsible. A human being at fault when something goes wrong does not improve the outcome.

[bobnikon]

How do you feel about structures or bridges “designed” by an “AI?”

Ambivalent. Are you concerned the software will make more mistakes than humans? Humans make a lot of mistakes.

Such a strange turn this thread took. Is everyone in some kind of mad panic because of some software hat tricks all of a sudden? Software has been swallowing industries for decades.

Yes, I am concerned software will make mistakes, but the worst part of it is that the responsibility for mistakes is removed.

Is that the case though? Truly asking. I can't imagine, at least any time soon, that any bridge would go to construction without a live engineer signing it off, or being involved throughout. I could be wrong. That is the responsibility and the conscience, for now at least. Though live engineers have also designed and signed off on some very sketchy designs.

A lot of the discussion is making it seem like "Skynet begins to learn rapidly and eventually becomes self-aware at 2:14 a.m., EDT, on August 29, 1997." At least we made it to 2024. I think we will persevere a little longer still. With what we have done to society and the planet, I wouldn't be surprised if the monkeys or the dolphins or the computers eventually decide we need to go. As Agent Smith said "Human beings are a disease, a cancer of this planet. You are a plague, and we are the cure."

Until then, everyone will keep eating up self driving cars and greater automation that makes our lives easier and our minds and bodies softer, because we like easy.

And back to the OP, interesting pieces, but not really up my alley.

[weeping minora]

Yes, I am concerned software will make mistakes,

Are you concerned humans will make mistakes?

The concern is understandable but keep in mind software is infinitely more testable and verifiable than humans are.

but the worst part of it is that the responsibility for mistakes is removed.

I’m not sure how to interpret this. The responsibility does not change. The company that builds the bridge is still responsible. A human being at fault when something goes wrong does not improve the outcome.

I'm more concerned that the accountability for those who have made mistakes is becoming much more lax than it should be. I see this happening within my workplace, and amongst criminals/LEO in striking observance, and at a much more alarming rate than the overall implicit negatives that AI poses at this particular point in time. I'm also worried that too many people will become too reliant on AI to perform, or hold accountability for, human responsibilities, rather than the potential for mistakes to be of overall heightened concern. These two points together could cause a bigger disaster than the overall Good vs. Evil of advanced technology, itself. If we lose the human element in the advancement of AI, we will be much more susceptible to negative outcomes for the future.

[RustyIron]

Such a strange turn this thread took. Is everyone in some kind of mad panic because of some software hat tricks all of a sudden?

I'm afraid you're just a lone man yelling into a hurricane. Good luck. I hope you succeed. People need something to panic about nowadays. Has it always been like this? I think it's because people no longer do things to occupy their time and develop their intellects. They sit around, their minds run away unchecked, and goofy notions are the result. I'm afraid the only solution is to address the lack of education in mathematics and science. Tik Tok, Snapple Chat, and--dare I say it--Internet forums are the breeding grounds of sloppy thinking.

[JayHenMac]

I can see a bunch of legal issues with AI. Patents and trademarks, etc.

sal

This is already a big issue in many industries including mine. The question is whether the output of the AI is the actual end product or just a tool/resource used to produce something. Where do designers get inspiration? Art? Nature? Force of function? Magazine images? AI generated images?

We run into issues with engineers using AI to develop code. Some argue that it's cheating, some argue that it's merely using better tools to produce a better product. Too many grey areas.

Yes, if someone rips off a patented product or design element, it's clearly a foul. What if they didn't but were simply using the AI to come up with ideas on which they base their work?

In my industry, ANY time AI is used as inspiration, or as a tool to develop code, or any solution, it has to be clearly and thoroughly documented. Even then the legal department can, and sometimes do, step in and kill a project. As they should.

[Naperville]

The fundamental question people should be asking is:
Is it net positive or net negative?

Just be honest with yourself when answering this.
I hardly see any positive.

It's a new fangled contraption so everyone wants to try it. As I said, I tried it and it is an above average summary device for topics you may be too busy to research yourself and don't care about the accuracy of the summary.

I'm far more worried about where we are headed with quantum computing destroying crypto, systems hacking and where AI is going say 5 to 10 years down the road. In a decade all of these technologies are going to upend everything and could cause wars.

Don't say that I did not warn you.

The “AI” are already capable of plenty more. It is hot topic in civil engineering industry. There are models being “trained” to make CAD drawings. How do you feel about structures or bridges “designed” by an “AI?”

Smaller AI systems that are trained on engineering data for one task are probably more accurate in their responses than OpenAI GPT-4, Bard/Gemini and Copilot. Anyone using AI for a public works project should check all results or they will be sued into oblivion when the wrong data is used.

LLM's only predict the next character or number in a sequence. To me it sounds a lot like gambling. I don't gamble that much and AI did not improve my life in any way when I used it. Over the last 3 to 6 months, I use it twice last week to respond on a Spyderco thread on knife prices, but I did not check the responses from Copilot. Nobody's life hung in the balance.

If you think AI is bad now, you haven't seen anything. When they figure out how to do more than predict the next character we could be in a heap of trouble.

[ChrisinHove]

I can see a bunch of legal issues with AI. Patents and trademarks, etc.

sal

This is already a big issue in many industries including mine. The question is whether the output of the AI is the actual end product or just a tool/resource used to produce something. Where do designers get inspiration? Art? Nature? Force of function? Magazine images? AI generated images?

We run into issues with engineers using AI to develop code. Some argue that it's cheating, some argue that it's merely using better tools to produce a better product. Too many grey areas.

Yes, if someone rips off a patented product or design element, it's clearly a foul. What if they didn't but were simply using the AI to come up with ideas on which they base their work?

In my industry, ANY time AI is used as inspiration, or as a tool to develop code, or any solution, it has to be clearly and thoroughly documented. Even then the legal department can, and sometimes do, step in and kill a project. As they should.

Interesting post. It stands to reason that it’s the responsibility of the user to ensure - and if necessary, prove in court - that their ai result hasn’t infringed copyright, patents or ip.