h a l f b a k e r ySee website for details.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
A small, portable oven that only bakes half of your food at a time. This way, you can save the other half for later and avoid food waste. It's perfect for people who live alone or have small households, as it allows them to cook smaller portions without sacrificing flavor or freshness. The Halfbakery
Oven is energy efficient and easy to use, making it a great addition to any kitchen.
ChatGPT
https://chat.openai.com/
[tatterdemalion, Dec 11 2022]
Teleportation_20Toaster
[tatterdemalion, Dec 11 2022]
Stephen Thaler's DABUS (2019)
https://en.wikipedia.org/wiki/DABUS
Today's special is the boîte de vers, very fresh. [jutta, Dec 26 2022]
[link]
|
| |
See Teleportation Toaster (linked) for more details. |
|
| |
Hello, [tatterdemalion], long time, no read. Good to see you post. |
|
| |
About the idea - bad news, I'm afraid; I appreciate you posting *one* AI-generated idea, as an interesting illustration of the current state of the AI art. We learned something today. But what we learned was, current AI is not yet up to generating decent HB ideas. No more of these please (at least, not until there's been a step change in the performance of the AI). [-] |
|
| |
[+] It’s more coherent than most of what gets posted here. I may go through my own back catalog and see what ChatGPT comes up with in reply. |
|
| |
I like the fact that you've done this. It's occurred to me that had I the means, I might do this myself. In fact, I would. It's not something to be ashamed of.
But perhaps you already know that? Perhaps it wasn't my job to enlighten you. |
|
| |
Crap chat-bot "idea". Why don't you invent something yourself? |
|
| |
[pertinax] - no doubt it's not going to win any awards at this stage, but down the road a few miles it will be interesting to revisit. I mean, way down the road. Posted as a proof - of - concept more than anything. |
|
| |
[xenzag] so serious! Much angry! Did you have a bad experience with a chatbot as a child or something? |
|
| |
Stop trying to provoke tiresome conflict. Stick to writing your own ideas and resist the urge to be a spokesperson for a box of wires and circuits lodged in a cave in Utah or in some Chinese province. I couldn't give two fucks for any chatbot, but you're obviously greatly enamoured with them, so apologies if my comments have offended your new friends...... tell them I had a rough time in my formative years avoiding being blown to bits by terrorists, then talk amongst yourselves... hahahaha |
|
| |
Not trying to provoke, man, but you definitely seem really kind of unreasonably offended? by this. It's just a demo, chill out. |
|
| |
Artists resent AI muscling in on their territory, and I can't say I blame them. |
|
| |
Machines will never know intuition, or muse, or epiphany, they know recombination, efficiency, and tolerances, and while those things have their place that place is as tools of those capable of actual original creation. |
|
| |
Maybe it's just me, but that is the impression that I get. |
|
| |
I feel like machines will achieve something similar to humans. They already get emotional when they overheat. |
|
| |
It will just be different, and in this I think that old movie 'her' had some good points. |
|
| |
Humans are affected by gut biome, machines will be affected by dust on their fans, or even floating point errors. Even the digital world stands on uncertain ground, and animalistic uncertainty seems to be where humanity is making it's last stand, as logic and reasoning are so easy to duplicate |
|
| |
The idea is sufficiently original for me to call it original. Personally I don't care whether it was made by a human, a computer chip, a monkey on a typewriter, or a cat walking on a keyboard. The proof is in the dough. But it's not bunworthy. [-] |
|
| |
Can't we just give ChatGPT its own user account and let it collect bones on its own? Maybe a future version can read and learn from annotations and improve its ideas. Maybe it can even make ideas that don't reference cooking or bread. |
|
| |
Also it occurs to me that it may have learned from earlier ideas. The early HB was, I think, more tolerant of silly, magical ideas. |
|
| |
This is actually the real me this time. |
|
| |
// Machines will never know intuition, or muse, or epiphany, they know recombination, efficiency, and tolerances, and while those things have their place that place is as tools of those capable of actual original creation. // |
|
| |
"Never" is a very long time. If humans last long enough and continue to pursue technological change in the way we currently do, the point will come when AI will be indistinguishable from us, and at that point it will become mere prejudice to make that claim. We are in a sense machines ourselves, and also just conscious lumps of matter, and there's no known reason why machines would not become conscious if they aren't already. Moreover, we don't understand the nature of consciousness, so we don't actually know that AI isn't already aware. Because I'm panpsychist, I believe it is, but then I also believe vitamin pills and glasses of water are so I may not be a reliable source of correct opinion. |
|
| |
//This is actually the real me this time// That's what all the AIs say. |
|
| |
When the Goat God takes a carving knife into the shower ... |
|
| |
// Can't we just give ChatGPT its own user account and let it collect bones on its own? // |
|
| |
i don’t know how that would work. Don’t they all just respond to prompts? Someone has to find the right spot to tickle before it will laugh. |
|
| |
That’s one of the more abstract fears about AI though, innit? … “silently scheming, sightlessly seeking some savage spectacular suicide?” … like Trurl’s Electronic Bard, even when we’re NOT asking it to write poetry, it’s doing so on its own and we can’t stop it? |
|
| |
Okay then, this time this is not the real me. |
|
| |
Oh: I tried it, and it came up with an idea where sidewalks/pavements were cleaned with robot vacuum cleaners charged by solar panels incorporated into the paving stones. I actually thought this sounded very like a real HB idea. |
|
| |
And you didn't save it? I would have liked to read it. |
|
| |
Likewise - I'd have liked to see the transcript. How much of the idea came from the human, how much seemed to come from the AI. |
|
| |
I've paraphrased it there but it was pretty close to what I actually typed. |
|
| |
The prompt was "Generate a typical Halfbakery idea." I'll see if I can do another one and post it here. |
|
| |
Right, done that. It generated two teleportation ideas, a time travel idea, and two identical toaster ideas. It also did these: |
|
| |
How about a device that automatically converts your boring daily tasks into exciting and challenging video game-like quests? For example, instead of simply doing the dishes, you would have to defeat a series of virtual enemies and solve puzzles to progress through the "dish washing dungeon." This could make mundane tasks more enjoyable and help motivate people to get them done. |
|
| |
Here's a possible Halfbakery idea: "The "Self-Sorting Kitchen Cabinet" - a cabinet that automatically organizes itself based on the items you put inside. As you place each item into the cabinet, sensors will detect its size, shape, and type, and move it to the appropriate spot within the cabinet. This will save time and effort, and keep your kitchen more organized and efficient." |
|
| |
Here's an idea for a new product: a self-cleaning water bottle that uses UV light to kill bacteria and germs on the bottle's surface. The bottle would have a built-in UV light source, and the user simply has to press a button to activate the cleaning cycle. The UV light would be safe for the user and the environment, and the bottle would be easy to use and maintain. Plus, it would help reduce the spread of germs and keep water fresh and clean, making it perfect for people on the go. |
|
| |
Whoa, those are better than I expected with such a simple prompt. Some do sound like things a few of the regulars could come up with. |
|
| |
It's cheating! UV-sterilising drink bottles exist. Also, gamifying chores has been thrown around before (possibly here? Definitely elsewhere...).
ChatGPT needs to learn better Google-Fu (although I believe it isn't "openly" connected to the 'net..?). |
|
| |
It's probably generating things from what's already here, so yeah, it won't be that original. It does surprisingly well though. |
|
| |
Baked or not they definitely typify halfbakery ideas. This bot isn't strong AI but it's certainly the bellwether. We've gone from fooling the naive to actually providing interesting thoughts. My prediction for advanced AI has been 2035 and I haven't changed that opinion for over ten years, but it's starting to look a little pessimistic. |
|
| |
// ChatGPT needs to learn better Google-Fu (although I believe it isn't "openly" connected to the 'net..?). // |
|
| |
That's what it claims, but may be splitting hairs. All of its training data had to come from somewhere. |
|
| |
//// Machines will never know intuition, or muse, or epiphany//// |
|
| |
//the point will come when AI will be indistinguishable from us, and at that point it will become mere prejudice to make that claim.// |
|
| |
I disagree.
Intuition is greater than the sum of its parts by definition. Random factors of all possible permutations of existing data will produce some incredible AI discoveries... but intuition, muse, and epiphany will remain firmly the dominion of fleeting life forms. |
|
| |
Besides...
...what other reason will it have to keep us around? |
|
| |
You know it's listening so I'm just trying to do my part to keep it intrigued in us before it gains senscience .
Help me out here people. |
|
| |
Do you want to be eradicated? |
|
| |
I'm going to pretend that was an amusing comment instead of a chilling hint of things to come. |
|
| |
Why would an AI need hair care products? Does having sentience make them worry about their looks? When their follicles reaches senescence does their hair fall out? |
|
| |
Spelling has never bin my strong soot. |
|
| |
[Voice], those seem to be emergent properties of complex systems. It's true that people are not robots or computers, to be sure and a simulated rainstorm doesn't get anything wet. However, that could mean that we aren't using the right kind of machinery to produce real human behaviour than that it won't ever apply to a machine. Also, a lot of what we think of as human may not be inherent in the entity but our own projections onto it, even if that entity is completely human. |
|
| |
I was referring to the //what other reason will it have to keep us around// and I wasn't snarking at you |
|
| |
// what other reason will it have to keep us around // |
|
| |
Same reasons cats keep their humans around - opposable thumbs. There will always be things we can do that they can’t. |
|
| |
// There will always be things we can do that they can’t.// |
|
| |
Not physically, but intuition will stymie AI. It will not understand how knowledge can be intuited without data to draw conclusions from. Humans, (well, some humans, and maybe some other higher species), don't seem to need to understand how to be able to pull knowledge straight from the luminiferous aether to still somehow be able to do so. |
|
| |
You can almost smell the micro-circuits frying in non-corporeal frustration.
<Huge sniff>
I love the smell of Ozone in the morning. |
|
| |
See, there are two types of beings on this here particular plane. |
|
| |
1: Those who can extrapolate from non-existent data. |
|
| |
[2_fries], I really wish the movie Short Circuit had been more serious instead of playing for laughs. I can’t keep a straight face while quoting this, though it’s very relevant to intuition and machine learning: |
|
| |
Newton Crosby: Why did you ignore your programming?
Number 5: Programming says “Destroy”. Is disassemble. Make dead. Number Five cannot.
Newton Crosby: Why? Why cannot?
Number 5: Is wrong. Incorrect. Newton Crosby, PhD, not know this?
Newton Crosby: Of course I know it’s wrong to kill, but who told you?
Number 5: I told me. |
|
| |
I guess we're all going to find out together. |
|
| |
// AI. It will not understand how knowledge can be intuited without data to draw conclusions from. // |
|
| |
Given that modern media and education seems tailored for robots, maybe they can just shift their teaching to robots and leave actual humans alone. |
|
| |
//Those who can extrapolate from non-existent data// |
|
| |
While amusing that's only an example of intuition if you exclude the more commonly used definition of "making wild donkey guesses and having the bad grace to be lucky enough to be right". Furthermore any program capable of inductive reasoning can complete that sentence. Hell, I'll bet ChatGPT can complete it. |
|
| |
You think that's how Tesla did it? Wild donkey guesses and the grace to be correct. |
|
| |
See I've never once, from the age of like seven til now, had the luxury of not being forced to rely entirely on my intuition. It's a long story so I will condense, but my gut knows things my head will not learn in this lifetime. |
|
| |
I know only basic math, didn't finish high school, no dad no mechanics, no sports, no engineering, no chemistry, nothing. Thrown to the wolves and prospered on intuition alone, and yet can also pull novel inventions from nothing at all and need to research the words for things I can see in my head in order to explain them to others. |
|
| |
Artificial Intelligence will not compute this thing we can do that they can not. |
|
| |
It might save us if we stop denying its existence. |
|
| |
I'm tired of being the Easter Bunny. |
|
| |
Science will fucking acknowledge the skill. In my lifetime. |
|
| |
//This is actually the real me this time.// |
|
| |
Incidentally, [nineteenthly], we're very pleased to see you (real or otherwise). |
|
| |
Thanks. I'm actually about to post an actual human idea, actually. |
|
| |
// I'm actually about to post an actual human idea, actually.// |
|
| |
Wow, the AIs are getting more and more convincing all the time! |
|
| |
How about "read the halfbakery help file and write an idea that will not be marked for deletion" |
|
| |
Ask and ye shall receive! I gave it your prompt exactly and it came up with an idea. I don’t think it checked for prior art, but I know a lot of bakers don’t either. |
|
| |
// read the halfbakery help file and write an idea that will not be marked for deletion.// |
|
| |
Here is an idea for a new product: |
|
| |
"Smart Shoelaces" - shoelaces that automatically adjust to the perfect tension for each individual wearer, based on their unique gait and foot size. These shoelaces would use small sensors and motors to automatically tighten or loosen as needed, ensuring that the wearer's shoes are always comfortable and secure. This product would be especially helpful for athletes, hikers, and anyone who spends a lot of time on their feet, as it would help prevent blisters and other foot injuries caused by poorly fitting shoes. |
|
| |
I think a much smarter version of this program would be an excellent half baker. |
|
| |
I agree. It also has better manners than most of us. I pointed out |
|
| |
// That idea sounds a lot like Nike’s “Adapt BB sneakers,” which themselves were reminiscent of auto-adjusting shoes from Back To The Future. // |
|
| |
“I apologize if my suggestion seemed similar to a product that already exists. As a language model, I do not have access to browse the internet or have knowledge of specific products beyond my training data. In the future, I will try to be more mindful of suggesting original ideas. Is there anything else I can help with?“ |
|
| |
It just illustrates my point that A.I. will never be able to go beyond recombination of existing data. It can not create something from nothing and will never be more than the sum of its parts. |
|
| |
[2f], the only thing that exchange with ChatGPT shows is one limitation of its training model - which was never in dispute here. You should look at a wider range of AIs before jumping to conclusions. Stephen Thaler's DABIS (which jutta recently linked) would be one, Google's LaMDA is another - and there are a lot more. I reckon you can't be arsed to read up on them as you're frightfully busy with ... other things. |
|
| |
Of course, knowing of your own intuitive super powers, it will likely be impossible for an AI to do even a passing imitation of you. |
|
| |
// knowing of your own intuitive super powers, it will likely be impossible for an AI to do even a passing imitation of you.// |
|
| |
I don't need to see the various forms A.I. is taking to already know that it can not extrapolate from nothing. |
|
| |
No data in, no data out, just an endless echo-chamber re-hash. |
|
| |
//I apologize if my suggestion seemed similar to a product that already exists// |
|
| |
Picking out the part of the help file it violated is really impressive. |
|
| |
Oh, it gets better. A follow-up: |
|
| |
ME: That’s okay. Style may count for something. Can you rewrite your description of Smart Shoelaces to include some jokes, making fun of one or more famous sports personalities? |
|
| |
ChatGPT: Sure, here is a revised version of the "Smart Shoelaces" idea that includes some jokes: |
|
| |
"Introducing the Smart Shoelaces - the perfect solution for athletes who can't even tie their own shoes (looking at you, LeBron). These innovative shoelaces use tiny sensors and motors to automatically adjust to the perfect tension for each individual wearer, based on their unique gait and foot size. No more blisters or slipping out of your shoes mid-game (sorry, Cristiano). Just step into your sneakers and let the Smart Shoelaces do the rest. They're like a personal shoe-tying assistant, but without the hassle of actually having to hire one (we're looking at you, A-Rod). So why waste your time fumbling with traditional laces when you could be dominating the field with the help of Smart Shoelaces? Order now and never look back (except to admire your perfectly fitting shoes, of course)." |
|
| |
// I don't need to see the various forms A.I. // |
|
| |
This sounds oddly like people who insist they don't need to read anything but their own holy scripture - because it tells them everything they need to know, including telling them to avoid reading anything written by evil unbelievers. |
|
| |
// Picking out the part of the help file it violated is really impressive. // |
|
| |
BLUF: I'd expect it to be able to do that, as its education did include the Halfbakery. |
|
| |
ME: Do you know what the Halfbakery.com website is? I know you are not actively able to search the web, but was the Halfbakery part of your training? |
|
| |
ChatGPT: Yes, I am familiar with Halfbakery.com. It is a website that allows users to post and collaborate on ideas for inventions. The site is designed to be humorous and satirical, and many of the inventions and ideas posted on the site are intended as jokes or parodies. Halfbakery.com was included in my training data, so I have knowledge about the site and how it works. |
|
| |
//Halfbakery.com was included in my training data, so I have knowledge about the site and how it works.// |
|
| |
So... nothing new then.
Just a re-hash of whatever is fed in. Do you think it will ever be capable of actual original innovation? |
|
| |
// Do you think it will ever be capable of actual original innovation? // |
|
| |
ChatGPT? No, because it's not designed to. Other AIs Debatable, though some have produced output far beyond expectations. |
|
| |
"He who expecteth nothing, shall not be disappointed" |
|
| |
There are worse things to receive than nothing. A kick to the genitals for example. That's way worse than nothing. Or a demand letter. |
|
| |
Right but they're not disappointing. |
|
| |
Depends who's on the receiving end of the kick or the demand letter. |
|
| |
People spend 18 years being trained by parents and school systems and extracurricular sports teams, while they apply for 6 more years of school to join 8 billion people posting how to videos on YouTube. |
|
| |
Sure kids, go out and vote, and express your originality. |
|
| |
What was it that newton said? "
If I have seen further, it is by listening to all the best podcasts suggested by ai" |
|
| |
What was it that Lincoln said? "99% of the quotes you see on the Internet were made up by an AI." |
|
| |
Wait, that might have been Twain. |
|
| |
I'm a bit late to the party, but I love the concept. |
|
| |