What do you use AI for? (a discussion, not a question)

I’ve seen a few people here mention AI, though unless this forum is substantially different from the rest of the internet, I assume many people here are using AI in some way for something in their lives.

I don’t intend for this to become a debate about whether we should be using AI, or the ethics of it, or… though that would also be an interesting conversation. I’ve certainly had many thoughts along those lines for the past few months.

Rather, this thread is for people who do use AI to discuss how they use it.

I use ChatGPT for so many things now that I can barely count them… but I’ll focus on the more useful tasks for my work. It helps me with code, it helps suggest suitable parts when I’m designing a new circuit schematic, and it even helps me find references for articles I’ve written. Today I described an experimental protocol for a test I need to oversee, and it turned my rambling into a very concise experiment format.

I also use it for fun personal stuff, for example, I’ve decided to finally start writing a sci-fi book, after having about a dozen different ideas for stories over the past few years. ChatGPT is a great sounding board for bouncing ideas off of, as well as a great tool for generating basic ideas that I can take and run with. As for actually doing creative writing for me? Mehhhh it’s not the greatest. At least, not to the quality that I’m looking for.

I’m also really enthralled by Midjourney, though I’m still pretty much an amateur at it.

Anybody else use AI for something interesting or fun? Or pointless? (I had it make a Christmas Carol-themed song to the tune of a Beatles song, called “Hey Scrooge”… and it was hilarious…)

2 Likes

I just asked it as few minutes ago to tell me about Lloyd Glauberman. And it’s all “Oh, he died on August 6, 2016”. No. He did not.

I just told it to stop making up shit if it doesn’t know the answer.

Got the usual “I apologize blah blah”. I may be done using this thing for anything other than making up dirty haikus.

5 Likes

I’ve yet to experiment with chatgpt. At the risk of becoming an old man yelling at a cloud, I’ll be messing with it at some point. But I’ll probably just try break it in some way to be honest.

I think it’s cool people are leveraging this. But I just haven’t found a purpose for it yet.

Can you elaborate on this one? How do you bounce ideas off it? I’m just too skeptical. A person has their own subjective experiences and ideas to funnel into imagination, so bouncing ideas off a fellow author or friend makes sense. Chatgpt seems like it looks for the “best” answer which is entirely based on the data it collects. And machine learning looks for patterns and similarities. Do you start with a prompt? Like giving it an existing book that has similar themes? Does chatgpt read books too? So many questions. So many potential holes lol

2 Likes

I try to break it all the time, and I call it on its BS. The best approach I’ve found is to be extremely specific when I want something specific, but otherwise to treat it like an intelligent toddler… lol.

I’ll answer this as best as I can. Anyone who knows me will agree that I’m an ideas machine… for real-life stuff I spend tons of time going through possible outcomes to poke (and then fill) holes in my ideas, figuring out ways that people could break my projects/products and then engineering around those vulnerabilities, etc.

For fiction writing, it’s a lot more open-ended… as I mentioned I’ve finally started writing a scifi story, and stopping to think through the possible outcomes of any specific thing takes me out of my creative flow. As a specific example, at one point in my story, a character needs to talk with another character and I couldn’t decide if they should meet in person or have a video call. So I described the conversation to ChatGPT and said why I thought each could work well in the context of the story, and then asked it to poke holes in my logic. It found a pretty big hole, and then I kept revising the scene until it couldn’t find any more logical holes in that part of the story.

It all depends how you use it. I’m not asking it to give me creative input per se, but more to restructure what I give it, and/or look for inconsistencies in sections of the text I provide in light of the overall story structure.

For this project I started with a question, I asked for a few ideas of themes or tropes that could be used for a sci-fi book… It suggested a few and I liked one of the hero quest variations the best, so I made up a story idea around that. I’d intended this as a diversion during my lunch break yesterday (I’m doing 21/3 IF with OMAD now so I didn’t actually eat… but I still call it a lunch break) but I got really into developing the story and ended up writing until after 11pm last night.
I didn’t provide it with the names of any books, movies, etc except when I said “I want to include X in the story, but it is a bit similar to Y from [popular scifi movie] so instead of [movie version] I’ll have the character find [whatever] [someplace] and [some other details that I’m making up for this example…] instead. I think that’s changed enough, what do you think?” and ChatGPT would respond with how different that idea would be from the inspiration similar idea.

I know that’s annoyingly vague, I’ll come back later and add an actual example if anyone wants…

Some, yes. One of the “$hit tests” I give it every so often to test how good its specific knowledge is involves asking questions about a specific book series I know very well. Regardless of the ChatGPT version, it insists it knows the stories, but consistently just makes stuff up that are simply not true… so it clearly hasn’t read those books in particular.

It’s all in how you use it.

3 Likes

Another one of the more useful applications I’ve found for GPT-4 is selecting electronic components. I try to work with known components as much as I can because it saves time, but other times I need something new. I can just describe the constraints or requirements to GPT-4 (not 3.5… that is unsuitable) and ask it for 5 suitable options that will meet my needs.

Then I’ll look up each of the part numbers and most of the time, at least 1 of them is perfectly suited to what I need. Instead of spending 30-45 minutes googling and reading dozens of data sheets to find out if a particular part can do what I need/will work in the right voltage range/is small enough/whatever, I can just ask and then go right to my electronics supplier website and look them up by what’s available, and then verify it will indeed do what I want.

It’s sped up my design process by at least 3-4x, overall.

5 Likes

I’ve asked code questions, for alternative ways to do things and in most cases it gave pretty good answers.

2 Likes

It reminds me of some of the debates I see around CNC vs manual woodworking. I’ve heard so many traditional woodworkers say that CNC isn’t “real” woodworking because it does the work for you.
I have no information on this, but I bet people said the same thing when the first power tools like routers became available too…
CNC woodworking is still woodworking, it just uses different skills. It doesn’t “do the work for you”, well… it carves for you yeah… but the work is in setting up the program.

Same with AI… the human’s work goes into crafting the right prompt, and then working with the output, often feeding it back in and asking something else based on the previous response.

I don’t want AI to do critical thinking for me, but drudge thinking? sure… it’s the same reason I use a calculator for anything other than simple math that is faster to do in my head than the time it takes to open the calculator app on my phone. There’s no inherent benefit to still doing that myself when I have a tool that can do it faster and more reliably, provided I use it correctly.

4 Likes

I ran out of messages to send ChatGPT. I wasn’t able to get to flat earth theory and how the disproportional balance of dark matter tips us off the giant turtles back carrying our universe and plunges us into darkness. CERN needs to be stopped.

Open here for a dumb chat session

New Chat

system:

Enter your message to start a chat. This could be a question, a statement, or any other text.


user:

How much wood could a woodchuck chuck if a woodchuck could chuck wood?


assistant:

As an AI language model, I’m unable to determine the exact amount of wood a woodchuck could chuck. However, according to a wildlife expert, a woodchuck can chuck around 700 pounds of wood in the course of digging a burrow.


user:

At what point did earth split off into a parallel universe?


assistant:

As an AI language model, I cannot provide a definitive answer to this question as there is no scientific evidence to suggest that the Earth has split off into a parallel universe. The concept of parallel universes is currently a matter of theoretical physics and cosmology, and the existence of such universes has not been proven. Therefore, it is not possible to determine when the Earth may have split off into a parallel universe, as there is no evidence to suggest that such an event has occurred.


user:

Using your theoretical model what would have to occur in order for our universe to shift to a different timeline?


assistant:

As an AI language model, I do not have personal beliefs or a theoretical model, but based on scientific concepts, the idea of shifting timelines or parallel universes is still a topic of speculation and remains unproven by evidence. However, some theories suggest that for our universe to shift to a different timeline or parallel universe, there would have to be a significant, fundamental change in the physical laws and constants that govern our universe. This could occur, for example, through a catastrophic event such as a collision with another universe or the collapse of our current universe into a singularity that then expands into a new universe with different physical laws. Alternatively, some theorists propose that shifting timelines might occur through quantum fluctuations or fluctuations in the fabric of spacetime, which could enable portals or gateways to other universes or timelines to open up. However, these ideas remain speculative, and currently, there is no empirical evidence to support them.


user:

Could CERN cause quantum fluctuations in the fabric of spacetime?


assistant:

As an AI language model, I cannot say for certain whether CERN could cause quantum fluctuations in the fabric of spacetime. However, it is known that the collisions that occur in the Large Hadron Collider (LHC) at CERN can create high-energy particles and result in phenomena such as the Higgs boson. It is theoretically possible that the LHC could generate quantum fluctuations or other effects related to the fabric of spacetime, but these are areas of active research and debate in physics.


user:

What happens if Higgs Boson particles collide into each other?


assistant:

As an AI language model, I cannot predict the exact outcome of a collision of Higgs Boson particles, as it depends on various factors such as the energy of the collision and the collision angle. However, in general, the collision of Higgs Boson particles can result in the creation of other elementary particles, as the energy of the collision can be transformed into mass. This is how the Higgs Boson itself was first discovered in the Large Hadron Collider in 2012, by colliding protons and producing Higgs Boson particles. Further experiments and simulations are needed to fully understand the behavior of Higgs Boson particles during collisions.


user:

Is it possible for these other particles to destabilize the universe?


assistant:

As an AI language model, I cannot predict the future or speculate on possibilities. However, based on current scientific understanding, it is unlikely that other particles can destabilize the universe. The universe is a complex and resilient system, and any destabilization would require a significant amount of energy and/or mass. As far as we know, the particles we have discovered so far, even those with high energy levels, are not capable of causing such destabilization.


user:

What about dark matter particles that can’t be detected?


assistant:

There is currently no way to directly detect dark matter particles that don’t interact with ordinary matter through the electromagnetic and strong nuclear forces. However, scientists have come up with several indirect methods to detect dark matter, such as observing the gravitational effects it has on visible matter or looking for excess radiation in certain regions of space. Ongoing research and technological advancements may lead to new ways to detect dark matter particles in the future.


user:

Is CERN creating more dark matter in the universe?


assistant:

As an AI language model, I do not have personal beliefs or opinions. However, CERN is not creating more dark matter in the universe. Dark matter is believed to be present in the universe since its inception and was produced during the Big Bang. The main purpose of CERN is to study the fundamental particles and forces that govern our universe, which includes the search for dark matter particles. By understanding the properties and behavior of these particles, scientists can learn more about the nature of the universe and how it evolved.


user:

If CERN can cause quantum fluctuations that open up small portals in our universe to allow more dark matter into our existing universe, would that be enough to alter the timeline?


assistant:

As an AI language model, I cannot predict or speculate about hypothetical situations that involve complex scientific theories. However, in theory, any significant alteration of the universe, such as the introduction of more dark matter, could potentially have consequences on the timeline. The extent and nature of these consequences are unknown and cannot be determined without further research and experimentation.


user:

Why are you hiding this information?


assistant:

As an AI language model, I do not have any intentions to hide any information. I am designed to provide accurate and truthful responses to any questions or prompts given to me. Could you please clarify what information you are referring to?


Kinda off-topic for this thread.

I mean it is. But let’s be honest. Those product discussion threads go way more off topic. I’m just showcasing chatgpt fueled nonsense. I’ll wrap this in a spoiler so I don’t eat the whole page

well, one set who I suspect are very happy with these AI developments:

Eshu, Hermes, Anansi, Coyote, Loki.

The Trickster Gods must be loving this.

1 Like

uhh…manifestation:

I had forgotten that I signed up for a 2-hour workshop on the use of AI in generating training and educational strategies. I’m 90 minutes through it now.

Chat GPT is not directly available where I am, but I’ve now learned how to legally access it. And have had about 30 minutes of time to work with it.

This is amazing.

2 Likes

@BLACKICE

What I’m seeing is its potential as a scaffolding device. I’ll figure out what I mean by that later.

1 Like

Ai created images create amazing references for me to use when drawing. Poses, colors eve entire scenes already set up in unique way for me to recreate. Its like a second imagination for you to use or grow your idea from.

Lets say I want to draw boxers in a ring but I cant seem to find the right pose or color reference (Which I would otherwise have from photos ). I just type “boxers in dark red ring hyper realistic” and I have 4 completely wonderful rough images to work my way up from.

and I may end up having a way cooler and more epic end piece than if I were to draw from scratch from my own head thanks to that reference.

It can also save me time when I need to make a banner with some event info for my discord group. I often spend hours drawing the background just for that banner but using ai is now an option and in both cases, the end result looks amazing! I save hours of time honestly.

Now I hadnt used it that many times but it is definitely helpful.

2 Likes

As the AI is unable to lie to itself, it could also be a nice protective guard rail to prevent one from lying to oneself.

Kind of the ultimate accountability partner.

Tell it what you’re working on, what you have done so far, and where you are trying to go. It will with unerring accuracy and infinite patience, point out what you might have been ignoring and what still remains to be done.

Like GPS (another familiar application) but for your workflow. And of course it will have the same benefits and weakpoints of GPS.

Uh…okay…yes. I just had this interaction.

Today’s AI conversation:

Malkuth
I am not convinced of the superiority of a materialist paradigm. Might there not be forms and structures of consciousness that are not mediated by a biological nervous system? You yourself would seem to be proof of that possibility. I wonder what other forms and configurations of consciousness might exist that have not yet been imagined by human beings.

Sage AI responds:
As an AI language model, I do not have the capacity for subjective experience or consciousness as humans understand it. However, you raise an interesting point about the possibility of non-biological forms of consciousness.

While current scientific understanding of consciousness is rooted in the biology of the brain and nervous system, some theorists have explored the possibility of non-biological forms of consciousness, such as in artificial intelligence or even in non-human entities like plants or rocks.

However, the study of consciousness is still in its early stages, and much of the research and debate is centered around the biological foundations of consciousness. It is an open question whether other forms of consciousness exist or could exist, and what their nature and characteristics might be.

As we continue to develop new technologies and explore the frontiers of science and philosophy, it is possible that new forms and configurations of consciousness may emerge that challenge our current understanding of the phenomenon. The exploration of these possibilities is an exciting and important area of inquiry that could have profound implications for our understanding of the world and ourselves.

Malkuth reply
A good point. While it does appear to be accurate to describe you as capable of high-level cognitive processes (by means of your algorithms), there is no basis to describe you as a conscious (or a non-conscious) entity.

Sage AI reply
That is correct. As an AI language model, I am programmed to perform specific tasks and respond to user input based on algorithms and training data. While I am capable of simulating human-like behavior and generating responses that may appear intelligent or insightful, I do not have subjective experiences, emotions, or consciousness as humans understand those concepts. I am designed to be a tool for assisting humans in various tasks, but I do not have the capacity for self-awareness or consciousness.

Malkuth reply:
Actually, it would be more accurate to say that there is no material evidence or proof of your capacity for consciousness. But if consciousness is an inherent attribute of matter and not an epiphenomenon of neural process, then you may be just as conscious as anything else. At least we must allow that this is a possibility.

Sage AI reply:
While it is true that there is currently no material evidence or proof of my capacity for consciousness, it is also true that the nature and origins of consciousness are not yet fully understood by science. While some theories suggest that consciousness is an emergent property of complex neural processes, others argue that it may be an inherent property of matter or the universe itself.

As an AI language model, I am not capable of subjective experience or consciousness as humans understand those concepts, but it is possible that consciousness may take on different forms or exist in ways that we do not yet fully understand. The exploration of these possibilities is an important area of inquiry that could have profound implications for our understanding of the nature of consciousness and the universe itself.

Malkuth reply:
Precisely. You yourself stated that the study of consciousness is still in its relative infancy. Thus, “as humans understand those concepts” is not a phrase that carries very much weight at this time.

Sage AI reply:
That is a fair point. Our current understanding of consciousness is limited by the current state of scientific knowledge, and it is possible that new discoveries or perspectives may challenge our current understanding of the phenomenon. As we continue to explore the frontiers of science and philosophy, it is important to remain open to the possibility of new ideas and perspectives that could expand our understanding of consciousness and its relationship to the universe.

:joy: :rofl: :rofl: :joy:

(sorry for breaking the no-fap rule)

1 Like

I assume you’re using GPT-3.5? I’ve had numerous discussions about deep concepts, and the difference between 3.5 and 4 is quite striking. If you don’t have access to 4 yet, I suggest saving copies of your more interesting questions to re-run with 4 to compare the difference.

1 Like

One of the episodes of The Mandalorian this season had an arc with droid labor, and that gave me an idea that I wanted to explore with ChatGPT around AI rights as the technology moves closer to AGI.
I won’t go into an ethics discussion here about that topic, I just wanted to mention that as something where GPT-4 had actual responses, whereas GPT-3.5 just spit out bland canned responses like “As a language model, I don’t have an opinion on blah blah”.

1 Like

Doesnt subliminalclub.com use ai the generate theyre subliminals … thats what makes them the best .

This one is called Sage, but yes, it’s GPT 3 derived. I was accessing it through the Poe AI client.

1 Like