The logo was created for OctoSQL[0] and in the article you can find a lot of sample phrase-image combinations, as it describes the whole path (generation, variation, editing) I went down. Let me know what you think!
And btw. if you get access take a look at [1] before you start using it. A ton of useful bits and pieces for your phrases.
TLDR: DALL·E 2 is really cool, though takes quite a bit of work to arrive at a useful picture. Moreover, some types of images work better than others ("pencil sketch" is consistently awesome). As with programming, it's difficult to realize how much pieces you have to specify if you're not an artist - you don't know what you don't know.
How much did the credits for all this image generation cost you?
edit: found it in the article: "From a monetary perspective, I’ve spent 30 bucks for the whole thing (in the end I was generating 2-3 edits/variations per minute). In other words, not too much."
I also tried to make it generate an icon for a product and I managed to get it to show me interesting things, but never got to make it actually draw it as one. Do you remember which prompt resulted in this macOS-ish app shape?
thanks for the writeup. I looked at your other blog posts and I would like to read more about octosql (needs/specification, architecture, development strategies, challenges, DBMS protocols/interfaces/libraries).
And thank you for adding outer joins after I recently mentioned that they are missing!
There is no technical documentation available right now other than the readme. I'm planning to write it around September-December (together with a website for them).
You can share your email at jakub dot wit dot martin at gmail and I'll let you know when it's available.
My god it is so frustrating that I can't seem to get open ai access any time I have an idea for a project using dall e, gpt, for whatever reason, they won't approve my account.
I have to sit here and watch everyone else play with the fun "open" ai tools... company needs a name change if they're going to keep this up.
Never heard of that. So I looked it up and it seems a service completely based on discord? Both for the community and support (I presume) as well as accessing the service itself? There doesn't even seem to be any HTTP API. Weird :)
Yeah, it's a neat idea but it's extremely frustrating to use. A really really basic web frontend would make it so much more usable.
On the upside (for MidJourney), you're seeing a HUGE stream (they are hitting the 1 mil Discord members ceiling) of generated pictures and that kinda grows your appetite and you want to also try more and more prompts..
I think it is still in sort of a testing/early access phase. Discord only access is essentially a way of funneling everybody who wants to try it into their captured marketing venue without having to have one of those "give us your email" placeholder pages (also has a bonus "social" aspect where you're seeing what many of the other people who are using it are creating). The final product will presumably be more tailored and web-driven.
It's also an interesting way of balancing what I assume are high operational costs on the server-end by pawning off some of the hosting of assets onto Discord.
My significant other who had entered the queue several months ago as nothing more than "developer" got in last week. Don't give up! There's always Craiyon to scratch the itch in the meanwhile. You can start to play around with ways to write prompts, etc.
Afaik they are opening it up to a much wider audience in the recent weeks. I also got it just 2 days ago, and applied the same way as your SO, only providing "developer" and nothing else.
That makes me think of they actually target developers somehow. I also got in couple days ago providing just email and being soft developer.
I know at least one artist and one relatively popular youtuber (with over million subs) who applied to a waiting list much earlier than me and are still waiting.
If you look on the public Discord servers for DALLE/AI you will find active servers that take requests. They seem pretty active too and had all the services available.
This is kind of a weird take to me given that photoshop exists. (Tons of proto-computer vision algorithms in there, like basic convolutional filters.) I suspect you'd still get copyright if you modify it a bit somehow.
So, if a nascent company chooses to go down this same path of generating (or maybe _seeding_) their logo design with AI, have they essentially given up any ability to protect that logo going forward?
Logo's are generally protected by trademark rather than copyright. I don't think anything prevents you from using a generated logo with trademark. For example you could have a trademark on an orange square, even though you could never copyright it. In the same way a trademark could protect your product name even if it is a single English common noun, as long as it is distinctive in use within your trademark scope.
You can use it according to their license, but is it copyrightable is the question, and precedent so far seems to say no since a human didnt author it.
From a technical perspective, there has been a much larger adoption of diffusion models which make these types of generative art much more viable. There has also been breakthroughs in connecting images and text with models like CLIP. DALLE-2, Imagen, and a lot of other generative work are using these ideas to get even better results.
Big pretrained models are a huge contributing factor. Being able to take a model that already mostly knows language and a model that already mostly knows images and hook them up means you don’t need to do the entire end to end learning together.
This might not be a popular opinion, but I think all the work OP put in here is probably worth more than 50-100 bucks (which is the price of a logo on something like Fiverr). And to make things worse, the logo itself still needs to be cleaned up[1] as it's way too blurry to be seriously used as an app icon, etc.
I might have not been too clear about it in the article, so if I haven't, I agree!
All of this was just me finding a practical purpose to go for while having fun with Dalle. If I was really serious about a logo, I would definitely go and pay an artist. Both for monetary, as well as esthetic, reasons.
Though as far as an app icon goes, I think it's actually sharp enough. It starts looking bad when you zoom in a bit.
Maybe this isn't what the previous poster meant, but sometimes I will say black & white when really I mean monochrome. Monochrome logos show up all over the place especially with icons for web apps. And they are good for printing on apparel, accessories, etc. I really doubt they are concerned about faxing
Wrong. And it has nothing to do with what kind of company you have. A logo should always degrade to 1-bit (line art) representation gracefully, so it can be used in or on all kinds of media. It could be physical objects, prints on hats, silhouettes on glass... not to mention being recognizable at all sizes.
You don't refute my point. In fact, you strengthen it by providing no evidence why this should be a requirement of modern logos for software companies. You list a bunch of things a logo should be useable for in your mind, otherwise its not a professional logo. However, you don't explain why it must "degrade to 1-bit" for those random things nor why the logos should support things like "silhouettes on glass". I can think of a handful of use cases but hardly a minimum requirement for a good logo for the majority of software companies.
I've run several different types of businesses and even those that required print work never required or even benefited from black and white, or even monochrome as another commenter mentioned. We _always_ had the means and preference for full color: emails, brochures, documents, websites, t-shirts—it didn't matter. There was _never_ a time we needed to degrade the logo so significantly. From talking with others that appears to be extremely common in modern businesses, especially software, since the majority of our presence and revenue stream is online, and not glass silhouettes in our office.
As I said, outside of a fairly narrow range of real world use cases, this comment is outdated: "Ignoring this issue is the mark of an amateur." If you have one of those rare use cases, check that box, but otherwise it shouldn't be the norm or a requirement.
Your point has been roundly refuted, with evidence that you yourself cited in your reply. Your limited imagination will limit what you do with logos. Enjoy.
The software used was Topaz Labs Sharpen AI. How they define "AI" I can't say for certain, but they're apparently using models so I'm assuming there's some kind of machine learning involved. Their software does a really good job on photos and videos well beyond what a standard sharpen filter does. The upscaling features are also pretty awesome. (no I don't work for them)
Jeremy Howard describes this as "Decrappification"[1]. This is one of the easiest deep learning models to train, in my opinion, as you can generate your own dataset easily. You just get good pictures for the target, programmatically make changes that make the image "crappy" for your source, and train until your network can convert from crappy to good. Then you pass it something it has never seen, and whabam, your picture is sharper than before.
This still doesn’t work well as a logo IMO, no amount the sharpening. It probably needs to get redrawn with a proper vector editor, the lines cleaned up and colors simplified
It’s a good first draft and something to give to a designer, but can’t stand by it’s own as a serious app logo
It's a good start, but it's more of an illustration than a logo to be honest. It should work as a single color (white, black), at small scale and in combination with your product name.
It would need to be turned into a vector to scale properly but I can think of other apps that have complex logos, especially on the MacOS ecosystem. Git Tower comes to mind.
Yes, the main usecase for DALL-E is probably for illustrations next to a story/blog. Logos are much harder to get right, and unsurprisingly DALL-E is not up to the task (yet).
I’ve had luck with similar things by being careful about my text prompt. Asking for tiny icon sized images also seems to clue it into the stylistic constraints of tiny icons (like what you mention).
With DALLE 2, I’ll pretty much never hire a graphic designer again to make any kind of logo. Good riddance, I’m sick and tired of their pretentious justifications for charging upwards of $500-$1k or more for simple logo designs.
Trivializing the work of people outside of one's profession while giving more importance to one's own is as old as the human civilization. There are plenty of people cursing software developers right now for pulling six digit salaries in exchange for typing some dumb text on a screen.
I speak under correction here, but you won't have teams; you will just do it yourself. Like cad/cam software got rid of a LOT of machine drafters and the designers just became responsible for outputting the finished drawings. You'll be the entire team.
Ah man, I do like what Dall-E is doing right now, i'm curious about the possibilities, even thinking of using it as a start for digital art and then manipulating the output in photoshop, i'm fairly good/creative at that manipulation side and I like what I can do there but I am a terrible artist, so this is good to get the starting point.
However as it gets better, even that won't be needed and I'm concerned then what that means for the average person, or for me trying to get my skills up so i can increase income, only for it to get wiped out by AI at some point in the future.
I think it makes much more sense for simple illustrations for articles, presentations and books ("pencil sketch" style). For logos, especially since you'd usually want simpler shapes, less detail, with a lot of readability, I'd go pay an artist if it was for a company I was building.
For logos you want specifically a design that will work well in black and white, and you want assets that are vector art. At the point that AIs can produce that it's worth revisiting for logos, but I'd bet that's probably more on a "many years from now" schedule.
Heh, and you don’t think we won’t train AIs specifically for drawing logos where not only can you specify features that you want but even the demographics you want it to appeal to based on mass collection of data.
It reminds be a bit of working as a director in a theater. You tell the actors what you want, and it's never just a "line reading". That's sort of the equivalent of just drawing it yourself, because you can't -- not just that you lack the expertise, but that you need them to do their thing with their body, and it has to be done their way or it looks fake.
So you end up using language that's sort of reminiscent of that, creating an emotional picture. It usually takes multiple passes to transfer the whole idea from your head to theirs.
I'm told that animation directors end up doing exactly the same thing. A digital model really can do what human actors can't. You could say "make that eyebrow curve 10% more" to an an animator. But it won't work unless you tell them why and what it means.
Mods: I see the title got the purpose of the logo edited out, but I think at least adding "a logo for my Open Source project" would be a much better title.
Incidentially you can ask DALL-E 2 for "vector art" and it'll comply, with good enough separation that it can be traced with something like Inkscape into true vectors.
You can also ask for "black and white vector art" to limit the color palette.
It's not as simple as "take a bitmap image and make it a vector". Yes sure, they'll vectorize it, but it'll look bad, through no fault of theirs. When creating a good looking vector image, you generally need to take into account it being a vector from the beginning of the design process.
Thanks for this post, it helped me tailor my own search queries. Because of your post, I was able to discover a whole new realm to DALLE-2. For some reason, repeating the same query parameter at the end yields some rather interesting results.
The first one looks like the Bacardi logo with a dragon instead of a bat and the second one looks like a Charmander. I think the second one is interesting because most art I see with baby dragons look more dragon-like and less salamander.
I'd wonder if that's an artifact of the source data, drilling down in the possibility space to be more like some subset that duplicates the image label- for example pulling tweets with body text and alt text.
Alternatively I guess it could just pull harder towards the prompt, idk.
I was going to comment that both look very much like what you'd find in an advanced beginner's deviantart portfolio...like, late high school-ish age, I woudl guess.
The second is more 'advanced' to me than the first, possessing an actual style, but neither is anything I would consider high quality enough to serve as a project/company/site/personal logo.
I think a lot of DALL-E 2 outputs fall into the category of "extremely impressive that a neural network made this" and also "not quite up to the standards of a human expert". Like if you show me an output and told me a machine made it, I'm absolutely fascinated, but if you showed me the same image and told me a human drew it, I'd just scroll past without a second thought. Even so, there are some applications for which being able to generate a pretty okay image for a few cents is a great deal - I use it for things like D&D character portraits.
Of course, DALL-E 2 is not the end of of text-to-image research - it'll be interesting to see where we are a year from now.
It is creating better images than the huge majority of people would do. Cheaply.
As you say, an expert can do far better.
But having something artistic created that well exceeds the average ability is gobsmakingly astonishing. And for quick blast variety generation, it is world class.
This automates 50% of a modern tech company, now you just need to automate the code generation, which seems already good enough to be on par with modern tech companies. Seems like a manage type can run his entire tech company himself now.
I’m fascinated by how much this is exactly like working with a human artist who doesn’t really understand the domain that you are wanting to represent with an image. Iterate, iterate, iterate.
It seems like the most valuable thing this could do is get some of that early exploration out of the way faster and easier than a human can do it, get to two or three concepts that feel like they’re in the neighborhood, and then let a human expert take over and turn it to something final quality. That’s pretty cool.
At the end of the article I also described a bit how I would see the evolution of such a tool, and it looks like we're thinking very similarly.
---
Though I think the real breakthrough will come when Dall-e gets 10-100x cheaper (and faster). I would then envision the following process of working with it (which is really just an optimization on top of what I’ve been doing now):
1. You write a phrase.
2. You are shown a hundred pictures for that phrase, preferably from very different regions of the latent space.
3. You select the ones best matching what you want.
4. Go back to 2, 4-5 times, getting better results every time.
5. Now you can write a phrase for what you would like to change (edit) and the original image would be used as the baseline. Go back to 2 until happy.
I see this happening in all areas. Everything would be prompt-driven.
Do you like this? What about this? You simply nod or reject the solutions that you don't want.
Pretty soon somebody's expertise and experience is not going to be enough to continue paying them what they used to get before this magic blackbox appeared.
One day enterprises will realize they can just outsource that expert who's been reduced to simply typing prompts and nodding yes or no.
I am worried that the middle class is rapidly disappearing. We will own nothing and be happy seems quite ominous. The question is then what field is safe from advancements in AI?
The only field I can think of is doctors, lawyers, executives, buy-side money managers. Even their jobs will be partially automated but it will be safe as long as they generate revenue.
>Pretty soon somebody's expertise and experience is not going to be enough to continue paying them what they used to get before this magic blackbox appeared.
I doubt it, because the process of thinking of phrases to feed dall-e is really the hard bit.
This is ok for a logo like this where it’s fair to say the base level expectation is not super creative. This logo is cool, but it doesn’t really stand out or make the product ver distinctive. If I am running a hobby or OS project that’s fine, but if I was investing a lot in sales/marketing then paying a real artist to make something interesting and novel is a rounding error.
> This logo is cool, but it doesn’t really stand out or make the product ver distinctive. If I am running a hobby or OS project that’s fine, but if I was investing a lot in sales/marketing then paying a real artist to make something interesting and novel is a rounding error.
Q: Are there really logos out there that are "interesting and novel" and that "stand out or make the product [..] distinctive"? Which ones?
EDIT: (perhaps more importantly) are there interesting, novel, distinctive logos that actually contribute to profitability?
tbf I think when it comes to big company branding it's the opposite.
A lot of GPT iterations of the design has left the article author with something which is quirkier than your average logo, but also looks like clipart and probably doesn't scale up or down well or work in monochrome. Which is fine for OSS. (He might get more users from blog traffic about using GPT-3 to design his logo than he ever could from any other logo anyway)
But when it comes to bigger companies, the design agency are the people that sit in meetings with execs persuading them that a well chosen font and a silhouette of a much simplified octopus will work much better ("but maybe the arms could interact with some of the letters etc etc, now lets discuss colours). The actual technical bit of drawing it is the bit that's already relatively cheaply and easily outsourced, and plenty of corporate logos are wordmarks that don't even need to be drawn...
But, if everyone's jobs are automated, nobody is making any money, so nobody has any money to pay doctors, lawyers, executives, money managers, etc. You would think that if these types were thinking rationally, they would be fighting to expand the middle class so more people can pay for their services.
In the past, eliminating humans from one set of jobs has been balanced by a new set of opportunities for humans in different jobs. Usually, the new jobs are more valuable.
That's not utopianism. The new jobs can't always be filled by the people kicked out of jobs. It really sucks to be them.
But it does mean that it's not irrational for people to want to automate other people's jobs. The net amount of stuff generated increases, rather than decreases.
This pattern may not last forever. There's already some thought that we've generated more than enough stuff to guarantee a decent standard of living to everybody (at least in the developed world) without working, and plenty more for luxuries if people choose to work. Even if we haven't reached it, we appear to be heading in that direction sooner rather than later.
That may cause a radical re-think at some point. And it won't be seriously delayed by making sure cartoonists have jobs.
> enough stuff to guarantee a decent standard of living to everybody
It's not a zero sum game. There's still growth in us. We'll go to space and expand 1000x more, the space has plenty of resources, and humans will have jobs working together with AI.
We'll have to automate childcare to make that happen. Otherwise, the birthrates of the rest of the world will follow the countries with the highest standards of living on a wild plunge into unsustainability.
Jobs are plentiful as long as wealth is well distributed.
In the past, fast automation has led to badly distributed wealth, and job loss. This situation has lasted until the unemployable people died off (yep, that was part of it), and enough wealth was redistributed through violent means.
Today we know better, and have really no reason to repeat the violent means of our previous revolutions. But it's really looking like the people in power want to repeat them.
> enough wealth was redistributed through violent means.
there were no instances of violent redistribution of wealth that ended better for the average person than before. Only that a different group of people ended up with wealth.
Automation makes stuff cheaper, even for people who didn't obtain any of the financial wealth via redistribution - because there's more than just financial wealth that get created with automation. New availability of services and goods (think internet of today - this is a wealth that couldn't have existed before, and one can benefit from it even if they are poor today).
You don’t need nodding or really any conscious reaction I think. It should be possible to have some camera directed at face hooked up to another AI that catches slight changes in pupil dilation or other changes imperceptible to naked eye and registers when something looks interesting to the user. You can then quickly show a stream of variations and pick the tagged ones and use them to improve the guesses. I imagine something like this might one day become a preferred way of interacting with computers/AI.
Doctors are very vulnerable. Most of dermatology is simple pattern recognition. I can easily see AI lawyers beating human lawyers in litigation, too. An AI lawyer will have read every single case and know the outcomes, and can fine tune arguments for specific parameters like which judge etc.
> Most of dermatology is simple pattern recognition.
I have a few qualms with this app:
1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
2. It doesn't actually replace a USB drive. Most people I know e-mail files to themselves or host them somewhere online to be able to perform presentations, but they still carry a USB drive in case there are connectivity problems. This does not solve the connectivity issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the service, is it reasonable to expect to make money off of this?
Just like in Star Trek. They really knew what the end goal was didn't they.
> enterprises will realize they can just outsource that expert who's been reduced to simply typing prompts and nodding yes or no
Tbf a program averaging the market for a fact gives better returns than most of the financial industry, yet they still exist. Even if we can automate something doesn't mean we will, usually for pointless emotional reasons.
But on the other hand it's hard to say if in a 100 years humans will still be employable in any practical capacity for literally anything.
This workflow reminds me of a generative art program from the early 1990s, but I just can't remember its name. It was a DOS or Windows program that had a very curvy, fluid GUI with different graphics sliders. It would show you some random tiles and you choose one to guide the algorithm's next generation of tiles.
I wonder if Kai Krause lurks here at HN. I'd love to know how he's doing. Apparently he's still living in his castle, which he bought around 1999 [0].
Some-when in the 00's I read an article about him that he was putting advanced networking stuff into the castle and had the intention to start something like a "think-tank" (doesn't really fit it, but I don't know what I'd call it) where he and others would hang around and code stuff.
I found the article [1] from July 2002, "Lord of the Castle Kai Krause presents Byteburg II".
> So that 's Kai Krause's long-cherished plan: Now the software guru has finally opened a center for founders and developers from the IT and software industry in Hemmersbach Castle near Cologne -- the Byteburg II
I really wonder what he's doing to these days. His plug-ins were legendary, as well as the User Interface for Bryce [2]
Your comment really intrigued me to google this interesting person I had never heard about before. This may well not be used to you, but Kai has a not-a-blog blog that I stumbled upon on here http://kai.sub.blue/en/sizemo.html.
Some really interesting reads. I especially appreciated his articles on the passing of Douglass Adams (apparently a close friend of his!) and Then vs Zen.
Love him or hate him (and I do both), Kai was all about cultivating his adulating cult of personality and dazzling everyone with his totally unique breathtakingly beautiful bespoke UIs! How can you possibly begrudge him and his fans of that simple pleasure? ;)
In the modest liner notes of one of the KPT CDROMS, Kai wrote a charming rambling story about how he was once passing through airport security, and the guard immediately recognized him as the User Interface Rock Star that he was: the guy who made Kai Power Tools and Power Goo and Bryce!
Kai's Power Goo - Classic '90s Funware! [LGR Retrospective]:
>Revisiting the mid 1990s to explore the world of gooey image manipulation from MetaTools! Kai Krause worked on some fantastically influential user interfaces too, so let's dive into all of it.
>"Now if you're like me, you must be thinking, ok, this is all well and good, sure, but who the heck is Kai? His name's on everything, so he must be special. OH HE IS! Say hello to Kai Krause. Embrace his gaze! He is an absolute legend in certain circles, not just for his software contributions, but his overall life story." [...]
>"... and now owns and resides in the 1000 year old tower near Rieneck Castle in Germany that he calls Byteburg. Oh, and along the way, he found time to work on software milestones like Poser, Bryce, Kai's Power Tools, and Kai's Super Goo, propagating what he called "Padded Cell" graphical interface design. "The interface is also, I call it the 'Padded Cell'. You just can't hurt yourself." -Kai
But all in all, it's a good thing for humanity that Kai said "Nein!" to Apple's offer to help them redesign their UI:
>read me first, Simon Jary, editor-in-chief, MacWorld, February 2000, page 5:
>When graphics guru Kai Krause was in his heyday, he once revealed to me that Apple had asked him to help redesign the Mac's interface. It was one of old Apple's very few pieces of good luck that Kai said "nein"
>At the time, Kai was king of the weird interface - Bryce, KPT and Goo were all decidedly odd, leaving users with lumps of spherical rock to swivel, and glowing orbs to fiddle with just to save a simple file. Kai's interface were fun, in a Crystal Maze kind of way. He did show me one possible interface, where the desktop metaphor was adapted to have more sophisticated layers - basically, it was the standard desktop but with no filing cabinet and all your folders and documents strewn over your screen as if you'd just turned on a fan to full blast and aimed it at your neatly stacked paperwork.
>Bruce “Tog” Tognazzini writes about Kansei Engineering:
>»Since the year A.D. 618 the Japanese have been creating beautiful Zen gardens, environments of harmony designed to instill in their users a sense of serenity and peace. […] Every rock and tree is thoughtfully placed in patterns that are at once random and yet teeming with order. Rocks are not just strewn about; they are carefully arranged in odd-numbered groupings and sunk into the ground to give the illusion of age and stability. Waterfalls are not simply lined with interesting rocks; they are tuned to create just the right b...
The model starts from 64x64 8bit RGB image of noise (random pixels) so technically 1 in 3_145_728 (64x64x256x3) but most will probably be very close to each other as the color difference won't be that much. The image is then further upsampled by two other models which will change some details, but shouldn't affect the general composition of an image.
Maybe I'm wrong, but with these diffusion models there is randomness in every sampling step too not just in the initialization and they can have 1000 steps to generate a single image.
Ah good point, this would introduce more variation if the initial noise is close, but if the initial noise is exactly the same it probably means it was initialized with the same seed and the rest of the generation will be the same since the random algorithms are deterministic.
Yes! It gives powerful tools for someone with a concept to get much closer to visualization of their idea.
DALL E 2 is like a low or no-code tool in that way.
The outcome may not be a "finished" product, especially as viewed by a professional designer (or web dev). However, its a heck of a lot better than a tersely written spec.
And in some cases, the product will work well enough to unblock the business, get customer feedback and generally keep things moving forward.
I think this is more powerful than a simple exploration tool. It took the author a long time to find a query format that generated logo-like images. Once they had that part down, they were quickly able to iterate on their query to find an image they liked. They were even able to fix part of the logo using the fill-in tool. I'm not sure why you'd bring a human into the mix, especially if you're on a budget.
Nah, people will leave out the professional. The same wild west grab whatever you can, steal, plunder to the detriment of artists, writers etc. And when the legislation arrives it will be already too late, accidentally.
Why should there be legislation? Do you want to restrict what people can do, just to force them to employ artists and writers? We could also forbid people from filling the gas tanks in their own cars, to protect the job of gas station attendant, but nobody wants to live in New Jersey.
you remember the concept of dumping, i.e., flooding a market with below cost product to drive out competing businesses? This is dumping for creatives.
editing: not that it's intentional, but these things will have the same effect; way too much product even for creative works. No one will be able to make money off the product but the tools.
Yeah, my first thought was "Ok, but you are going to need to involve a graphical artist to actually really make use of that logo". Like you probably want a vector version and you definitely need simplified versions for smaller sizes but then I stopped and realized how amazing this actually is. It "saved" (I know, it cost $30 but that's a steal for something like this) all the time and money you would have paid for iteration after iteration and let the author quickly hone in on what they wanted.
As someone who is incredibly terrible at graphic design but knows what they like this could be a game changer as iterations of this technology progress. I can imagine going further than images and having AI/ML generate full HTML layouts in this iterative way where you start to define your vision for a website or app even and it spits out ideas/concepts that you can "lock" parts of it you like and let it regenerate the rest.
I'm not downplaying designers role at all, I'd still go to one of them for the final design but to be able to wireframe using words/phrases and take a good idea of what I want would be amazing, especially for freelance/side-projects.
I think your art/design/craft is pretty good. Some people use pencils, some use Adobe products, you have gone out there and tried the new Dall-E medium.
Glad you thought out the usage, I am sure that when the novelty wears off that you will have that neat-as-octocat logo sorted out.
I appreciate that you appreciate the value that highly skilled designers bring to a product with their visual expertise.
However, I would like to see you A/B test the Dall E logo versus the winning designer logo. You could show odd IP addresses one logo and even addresses the other.
I think the designer would edge the robot for what you need (a logo), however, the proof is in the pudding and conversion rate.
Honestly though the hard part is the actual design which is already done here. Learning to vectorize a raster is something that can be done in a weekend with Inkscape, there's no reason to involve an actual graphics designer with this anymore.
Time value of money. The most optimal use of money and time would be getting the ML to iterate until you have the finished product, then get a designer to vectorise it and fix it up. That way you pay the designer for one iteration and spend all the time you would have spent iterating with the designer iterating with the ML model instead.
> Learning to vectorize a raster is something that can be done in a weekend with Inkscape, there's no reason to involve an actual graphics designer with this anymore.
If you lined up 100 resulting images, 99 from weekend beginners and 1 from an actual artist. I guarantee you that you would pick out the artist every time.
It might be simple to trace over an image but you are probably better getting an artist to spend 2 hours on it, it will most likely look better than 2 weeks of tracing.
I think you might be underestimating how much work goes into the last mile of a design. A lot of refinement work goes into typography in particular, a domain Dall-E isn’t yet proficient in at all.
Plus there is no reason why someone couldn't build a specialised AI model to do vectorisation and another to generate simplified versions of vectors.
People are already doing by combining DALL-E 2 with gfpgan for face restoration. So there may be a role in understanding how to combine these tools effectively.
An experienced human designer, right away, is going to ask how you want the logo to be used. That's going to have a major impact on how it's designed.
So yeah, this may be like working with a doodler, but, as the author intimated, this is far from an ideal experience in getting a professionally designed logo. This is more like "Hey, you, drawing nerd, make this thing."
Nevertheless, astonishing technology in its own right.
> unfortunately can’t do stuff like “give me the same entity as on the picture, but doing xyz”
That's my main gripe with DALL·E as well. This missing feature makes it impossible to use for stories where the same character goes through an adventure and is present in different settings, doing different things.
Although I don't know much about how DALL·E works, I have the feeling it shouldn't be too hard to add this possibility. That would make it so much better / more useful.
Looks good for ideation. Could potentially be more useful for an agency or creative professional building logos. They can make vector art off of promising mock-ups. Also most of the generated images need to be simplified for sake of a logo, but a professional can do this better.
image generators are cool, but there's no shortage of them (midjourney & running your own on collab) and dalle2 has nonsensical bans (why does "pepe" go against the content policy?)
Open-ai has nonsensical censorship. Dalle might be popular right now, but open-ai won't survive if they keep up their ridiculous attempts at trying to control culture. I've already got something running on collab, new models are coming out, and midjourney just got a v3 update that blows dalle2 out of the water.
I was disappointed to see this on their site. I had a pretty good idea of doing a sort-of online art installation by grabbing crime data near me from local web sources and having images automatically generate of those crimes, but unfortunately many of those crimes seem to be too violent for their filter.
If a neural network trained on data sucked off of the internet constantly draws frogs with a halo of swastikas when you ask for “Pepe” then maybe there’s a problem that needs solving there.
With all respect possible, you generated something that a professional will create for 20 minutes on a napkin (in the context of logo idea).
Maybe your perception of "logo" needs more reference points. For example, this gallery of classics in Brand Identity will be a good starting point(use the triangles on top to navigate):
https://www.joefino.com/logos_html/L01_Xpand.html
There is no doubt in my mind that the next iterations of neural networks will remove all "overpaid" and "overconfident" design professionals, that's why I adapted to the reality and moved to frontend development.
All of this with clear realization that everything humans can do for a production processes will be augmented and removed. The nasty "humans" always want to be paid, more and more. They want to have rights and privileges. What a hassle.:)
What an intelligent and educating response. My comment may come as salty, but if you make an effort to visit the linked gallery, maybe you will have more "fresh perspective":)
If someone wants to pay a lot of money to a professional to create their brand identity that option is always there.
If someone else just needs something simple and passable there is Dalle.
And I’m sure there is every option in between where someone can use Dalle as a starting point and pass it to a pro, or a pro would even use Dalle as a way to brainstorm options.
Dalle is a tool that has empowered everyone. It shouldn’t be seen from a stereotypical luddite perspective as in your first post.
Yep. Interesting times ahead of us:) How we will be able to tell the difference?
QR Code/Genetic sample government approved app for human verification?
And what when people are certain that the machines are better in everything, who will want to chat, listen to music or watch paintings from the "lame" humans, when the robots will be the ultimate solution for every human need?
As the tech stands today, mediocre artists, designers, writers and content creators are likely going to be replaced entirely with AI.
I imagine it would make it very easy to “seed” a website or a platform with initial “users” and content.
I also imagine it will be (and likely is already) being deployed to create the impression of popular support (or lack thereof) of a politician, business or policy.
This is most certainly already happening. I find it kind of annoying not to know with certainty whether or not I'm engaging with a Genuine Human(TM) or not.
I'm unsure if it's confirmation bias, but I find myself noticing weird abberations in online comments that don't seem to be ESL related. (edit: it's probably just mobile swipe typing at play)
Your "real question" ultimately resolves itself because the moment the novelty wears off (and it will happen very fast), nobody will be interested in chatting with "robots".
> With all respect possible, you generated something that a professional will create for 20 minutes on a napkin (in the context of logo idea).
I feel like I understand where you're coming from, but often the phrase I hear by experts (I even use this myself in my space) is, "Sure, it only took 20 minutes to do this wiring/write this code/draw this logo, but it took 5 years to know what to make." Sure, the results aren't what you'd get if you paid a professional logo designer, but if you can get close enough, it's really cutting out the X years training necessary to get to that point.
>it's really cutting out the X years training necessary to get to that point.
This is exactly my point. With repetition and solid design foundation comes the intuition what is the right direction towards the accomplishing of the given task.
Some will say the design is a subjective, I would argue that designers' role is to move towards objectivity and away from the idea of "personal taste".
That's why I give a link to the works of the master in this craft.
This is exactly the same argument with the Copilot case. Is it capable to give some "boilerplate" solution - yes. Is this solution mediocre at best - yes.
It is not even close to what Ironov does. More like a tech demo. Ironov outputs a complete brand book and it's interface is set up for exploration and logo refinement.
If you needed to play so much with words, then it will eventually become a specialized task killing any benefit of having the AI doing the work for you since we eventually we will need to resort to specialists to use the AI to get the result we expect.
On the bright side the result may be better, it may be easier to become and "AI usage specialist" than specializing in many different areas, the result may include many intermediate results that a specialist would find too much work to do and, with a bit a patience (like in the presented case), the task can still be done without the need of an "AI usage specialist".
Currently, I think the problem is an UI one. There should be an option to allow the user to do something like: "from the last drawing, just add this..." or "in the last drawing, change the color/size/style of this and that...". This would be probably enough to achieve what the author wanted in a much smaller number of iterations.
There is also on more thing: the costumer doesn't know exactly what he/she wants from the beginning. So, it is normal to have a few iterations until something pleasing is achieved.
Create a logo generator site, allow users to pick something very limited like industry/field from a dropdown or something, generate say 9 logos with AI generated text discriptions that fit this selection and remember which one the user picked and use that data to build a network that generates good text descriptions to feed into DALL-E 2 based on a singe item selected by the user.
> If you needed to play so much with words, then it will eventually become a specialized task killing any benefit of having the AI doing the work for you since we eventually we will need to resort to specialists to use the AI to get the result we expect.
This matches my view of the idea that AI will replace programmers. My value isn't in the typing and the syntax, it's in my ability to turn a spec into an internally consistent design by resolving conflicting instructions and clarifying edge cases; and sometimes in knowing what the user wants when they are unable to express it themselves.
Even if AI winds up writing all of the code, someone with the programmer mindset still needs to define the problem in a concrete manner. They'll always have a job as a "machine-talker."
I just got access today. Can’t wait to try it out.
We produce a lot of content and the biggest hurdle in graphic creation is the back and forth with the designer, plus the lag between writing, designing, and publishing. This would make it easy enough that the writer can include a prompt for the illustration right in the text itself.
More than the costs, I’m excited about the efficiency gains and smoother workflows.
I can't get anything good from DALLE-2. It seems so fcking stupid. Whatever I try, it gives me total BS, sometimes it just refuses to generate anything complaining about ToS violation.
385 comments
[ 2.0 ms ] story [ 282 ms ] threadThe logo was created for OctoSQL[0] and in the article you can find a lot of sample phrase-image combinations, as it describes the whole path (generation, variation, editing) I went down. Let me know what you think!
And btw. if you get access take a look at [1] before you start using it. A ton of useful bits and pieces for your phrases.
TLDR: DALL·E 2 is really cool, though takes quite a bit of work to arrive at a useful picture. Moreover, some types of images work better than others ("pencil sketch" is consistently awesome). As with programming, it's difficult to realize how much pieces you have to specify if you're not an artist - you don't know what you don't know.
[0]: https://github.com/cube2222/octosql
[1]: http://dallery.gallery/wp-content/uploads/2022/07/The-DALL%C...
edit: found it in the article: "From a monetary perspective, I’ve spent 30 bucks for the whole thing (in the end I was generating 2-3 edits/variations per minute). In other words, not too much."
It gets expensive fast.
https://jacobmartins.com/images/dalle2/DALL%C2%B7E%202022-08...
I didn't prompt anything specifically, it came after a line of variations from a definitely-not-icon-looking picture.
Though I'd try tags like "iOS icon".
thanks for the writeup. I looked at your other blog posts and I would like to read more about octosql (needs/specification, architecture, development strategies, challenges, DBMS protocols/interfaces/libraries).
And thank you for adding outer joins after I recently mentioned that they are missing!
There is no technical documentation available right now other than the readme. I'm planning to write it around September-December (together with a website for them).
You can share your email at jakub dot wit dot martin at gmail and I'll let you know when it's available.
I have to sit here and watch everyone else play with the fun "open" ai tools... company needs a name change if they're going to keep this up.
On the upside (for MidJourney), you're seeing a HUGE stream (they are hitting the 1 mil Discord members ceiling) of generated pictures and that kinda grows your appetite and you want to also try more and more prompts..
It's also an interesting way of balancing what I assume are high operational costs on the server-end by pawning off some of the hosting of assets onto Discord.
I know at least one artist and one relatively popular youtuber (with over million subs) who applied to a waiting list much earlier than me and are still waiting.
And concerning creating a logo with such tools: Is there any consensus on an eventual copyright of such works?
https://www.smithsonianmag.com/smart-news/us-copyright-offic...
that's not all there is to this though obviously
[1] https://raw.githubusercontent.com/cube2222/octosql/main/imag...
All of this was just me finding a practical purpose to go for while having fun with Dalle. If I was really serious about a logo, I would definitely go and pay an artist. Both for monetary, as well as esthetic, reasons.
Though as far as an app icon goes, I think it's actually sharp enough. It starts looking bad when you zoom in a bit.
In its current state it's not a viable logo because, for one thing, it won't look good in black & white.
That sounds like a concern that stopped being relevant for many software companies a decade ago at least.
These days app icons and hero images are more important than whether you can fax or print the logo.
Ignoring this issue is the mark of an amateur.
I've run several different types of businesses and even those that required print work never required or even benefited from black and white, or even monochrome as another commenter mentioned. We _always_ had the means and preference for full color: emails, brochures, documents, websites, t-shirts—it didn't matter. There was _never_ a time we needed to degrade the logo so significantly. From talking with others that appears to be extremely common in modern businesses, especially software, since the majority of our presence and revenue stream is online, and not glass silhouettes in our office.
As I said, outside of a fairly narrow range of real world use cases, this comment is outdated: "Ignoring this issue is the mark of an amateur." If you have one of those rare use cases, check that box, but otherwise it shouldn't be the norm or a requirement.
Seems to have been blurred after the fact. The version linked in the article before cropping looked fairly sharp: https://jacobmartins.com/images/dalle2/DALL%C2%B7E%202022-08...
Plus even that uncropped one is already jpeg'd, whereas DALL-E 2 downloads are pngs, so there should be an even sharper version.
Maybe in the US but not worldwide.
They need a black and white variation, different sizes, and the underlying component assets.
So Dalle2 might actually be able to provide that in the future as well.
But for now - it's going go give you an 'image' which you have to get an artist to then clean up int a proper logo with assets.
I'm playing with DallE-mini on hugging face and am generally unimpressed, I'm not sure if its' the same Dalle.
I tried the main DallE website sadly don't have an 'invite'.
https://imgur.com/a/m3hDMZq
The software used was Topaz Labs Sharpen AI. How they define "AI" I can't say for certain, but they're apparently using models so I'm assuming there's some kind of machine learning involved. Their software does a really good job on photos and videos well beyond what a standard sharpen filter does. The upscaling features are also pretty awesome. (no I don't work for them)
[1] - https://www.fast.ai/2019/05/03/decrappify/
It’s a good first draft and something to give to a designer, but can’t stand by it’s own as a serious app logo
However as it gets better, even that won't be needed and I'm concerned then what that means for the average person, or for me trying to get my skills up so i can increase income, only for it to get wiped out by AI at some point in the future.
I think it makes much more sense for simple illustrations for articles, presentations and books ("pencil sketch" style). For logos, especially since you'd usually want simpler shapes, less detail, with a lot of readability, I'd go pay an artist if it was for a company I was building.
It's like saying "steal from artists, but only the good ones"
I've always been fascinated by how artists abstract the core notion of an image. It's stunning to see a computer do that.
And indeed, seeing what Dalle will draw when telling it to visualize stuff like "data streams" was very interesting.
So you end up using language that's sort of reminiscent of that, creating an emotional picture. It usually takes multiple passes to transfer the whole idea from your head to theirs.
I'm told that animation directors end up doing exactly the same thing. A digital model really can do what human actors can't. You could say "make that eyebrow curve 10% more" to an an animator. But it won't work unless you tell them why and what it means.
You can also ask for "black and white vector art" to limit the color palette.
[1] https://labs.openai.com/s/x2UP0MEmj2qNnKWTbko8rrso
2. “cute baby dragon, logo, digital art, in a dark circle as the background”
[2] https://labs.openai.com/s/JmOXAqjpR2ctmraDxEkB7twF
Thanks for this post, it helped me tailor my own search queries. Because of your post, I was able to discover a whole new realm to DALLE-2. For some reason, repeating the same query parameter at the end yields some rather interesting results.
Something strange about DALL·E is that if you just type gibberish by pounding randomly on your keyboard, it will still "work", i.e., produce an image.
Alternatively I guess it could just pull harder towards the prompt, idk.
The second is more 'advanced' to me than the first, possessing an actual style, but neither is anything I would consider high quality enough to serve as a project/company/site/personal logo.
Of course, DALL-E 2 is not the end of of text-to-image research - it'll be interesting to see where we are a year from now.
As you say, an expert can do far better.
But having something artistic created that well exceeds the average ability is gobsmakingly astonishing. And for quick blast variety generation, it is world class.
It seems like the most valuable thing this could do is get some of that early exploration out of the way faster and easier than a human can do it, get to two or three concepts that feel like they’re in the neighborhood, and then let a human expert take over and turn it to something final quality. That’s pretty cool.
At the end of the article I also described a bit how I would see the evolution of such a tool, and it looks like we're thinking very similarly.
---
Though I think the real breakthrough will come when Dall-e gets 10-100x cheaper (and faster). I would then envision the following process of working with it (which is really just an optimization on top of what I’ve been doing now):
1. You write a phrase.
2. You are shown a hundred pictures for that phrase, preferably from very different regions of the latent space.
3. You select the ones best matching what you want.
4. Go back to 2, 4-5 times, getting better results every time.
5. Now you can write a phrase for what you would like to change (edit) and the original image would be used as the baseline. Go back to 2 until happy.
Do you like this? What about this? You simply nod or reject the solutions that you don't want.
Pretty soon somebody's expertise and experience is not going to be enough to continue paying them what they used to get before this magic blackbox appeared.
One day enterprises will realize they can just outsource that expert who's been reduced to simply typing prompts and nodding yes or no.
I am worried that the middle class is rapidly disappearing. We will own nothing and be happy seems quite ominous. The question is then what field is safe from advancements in AI?
The only field I can think of is doctors, lawyers, executives, buy-side money managers. Even their jobs will be partially automated but it will be safe as long as they generate revenue.
Every art director at an ad agency just shrieked!
This is ok for a logo like this where it’s fair to say the base level expectation is not super creative. This logo is cool, but it doesn’t really stand out or make the product ver distinctive. If I am running a hobby or OS project that’s fine, but if I was investing a lot in sales/marketing then paying a real artist to make something interesting and novel is a rounding error.
Q: Are there really logos out there that are "interesting and novel" and that "stand out or make the product [..] distinctive"? Which ones?
EDIT: (perhaps more importantly) are there interesting, novel, distinctive logos that actually contribute to profitability?
A lot of GPT iterations of the design has left the article author with something which is quirkier than your average logo, but also looks like clipart and probably doesn't scale up or down well or work in monochrome. Which is fine for OSS. (He might get more users from blog traffic about using GPT-3 to design his logo than he ever could from any other logo anyway)
But when it comes to bigger companies, the design agency are the people that sit in meetings with execs persuading them that a well chosen font and a silhouette of a much simplified octopus will work much better ("but maybe the arms could interact with some of the letters etc etc, now lets discuss colours). The actual technical bit of drawing it is the bit that's already relatively cheaply and easily outsourced, and plenty of corporate logos are wordmarks that don't even need to be drawn...
That's not utopianism. The new jobs can't always be filled by the people kicked out of jobs. It really sucks to be them.
But it does mean that it's not irrational for people to want to automate other people's jobs. The net amount of stuff generated increases, rather than decreases.
This pattern may not last forever. There's already some thought that we've generated more than enough stuff to guarantee a decent standard of living to everybody (at least in the developed world) without working, and plenty more for luxuries if people choose to work. Even if we haven't reached it, we appear to be heading in that direction sooner rather than later.
That may cause a radical re-think at some point. And it won't be seriously delayed by making sure cartoonists have jobs.
It's not a zero sum game. There's still growth in us. We'll go to space and expand 1000x more, the space has plenty of resources, and humans will have jobs working together with AI.
Q: Am I the only one thinking of Golgafrinchan Ark Fleet Ship B?
In the past, fast automation has led to badly distributed wealth, and job loss. This situation has lasted until the unemployable people died off (yep, that was part of it), and enough wealth was redistributed through violent means.
Today we know better, and have really no reason to repeat the violent means of our previous revolutions. But it's really looking like the people in power want to repeat them.
there were no instances of violent redistribution of wealth that ended better for the average person than before. Only that a different group of people ended up with wealth.
Automation makes stuff cheaper, even for people who didn't obtain any of the financial wealth via redistribution - because there's more than just financial wealth that get created with automation. New availability of services and goods (think internet of today - this is a wealth that couldn't have existed before, and one can benefit from it even if they are poor today).
I have a few qualms with this app:
1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
2. It doesn't actually replace a USB drive. Most people I know e-mail files to themselves or host them somewhere online to be able to perform presentations, but they still carry a USB drive in case there are connectivity problems. This does not solve the connectivity issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the service, is it reasonable to expect to make money off of this?
Edit: Ahh, it’s the Dropbox comment of HN fame. Never mind.
Just like in Star Trek. They really knew what the end goal was didn't they.
> enterprises will realize they can just outsource that expert who's been reduced to simply typing prompts and nodding yes or no
Tbf a program averaging the market for a fact gives better returns than most of the financial industry, yet they still exist. Even if we can automate something doesn't mean we will, usually for pointless emotional reasons.
But on the other hand it's hard to say if in a 100 years humans will still be employable in any practical capacity for literally anything.
Some-when in the 00's I read an article about him that he was putting advanced networking stuff into the castle and had the intention to start something like a "think-tank" (doesn't really fit it, but I don't know what I'd call it) where he and others would hang around and code stuff.
I found the article [1] from July 2002, "Lord of the Castle Kai Krause presents Byteburg II".
> So that 's Kai Krause's long-cherished plan: Now the software guru has finally opened a center for founders and developers from the IT and software industry in Hemmersbach Castle near Cologne -- the Byteburg II
I really wonder what he's doing to these days. His plug-ins were legendary, as well as the User Interface for Bryce [2]
[0] https://de.wikipedia.org/wiki/Burg_Rheineck
[1] https://www.heise.de/newsticker/meldung/Schlossherr-Kai-Krau...
[1, google translate] https://www-heise-de.translate.goog/newsticker/meldung/Schlo...
[2] https://en.wikipedia.org/wiki/Bryce_(software)
Some really interesting reads. I especially appreciated his articles on the passing of Douglass Adams (apparently a close friend of his!) and Then vs Zen.
Love him or hate him (and I do both), Kai was all about cultivating his adulating cult of personality and dazzling everyone with his totally unique breathtakingly beautiful bespoke UIs! How can you possibly begrudge him and his fans of that simple pleasure? ;)
In the modest liner notes of one of the KPT CDROMS, Kai wrote a charming rambling story about how he was once passing through airport security, and the guard immediately recognized him as the User Interface Rock Star that he was: the guy who made Kai Power Tools and Power Goo and Bryce!
Kai's Power Goo - Classic '90s Funware! [LGR Retrospective]:
https://www.youtube.com/watch?v=xt06OSIQ0PE&ab_channel=LGR
>Revisiting the mid 1990s to explore the world of gooey image manipulation from MetaTools! Kai Krause worked on some fantastically influential user interfaces too, so let's dive into all of it.
>"Now if you're like me, you must be thinking, ok, this is all well and good, sure, but who the heck is Kai? His name's on everything, so he must be special. OH HE IS! Say hello to Kai Krause. Embrace his gaze! He is an absolute legend in certain circles, not just for his software contributions, but his overall life story." [...]
>"... and now owns and resides in the 1000 year old tower near Rieneck Castle in Germany that he calls Byteburg. Oh, and along the way, he found time to work on software milestones like Poser, Bryce, Kai's Power Tools, and Kai's Super Goo, propagating what he called "Padded Cell" graphical interface design. "The interface is also, I call it the 'Padded Cell'. You just can't hurt yourself." -Kai
But all in all, it's a good thing for humanity that Kai said "Nein!" to Apple's offer to help them redesign their UI:
http://www.vintageapplemac.com/files/misc/MacWorld_UK_Feb_20...
>read me first, Simon Jary, editor-in-chief, MacWorld, February 2000, page 5:
>When graphics guru Kai Krause was in his heyday, he once revealed to me that Apple had asked him to help redesign the Mac's interface. It was one of old Apple's very few pieces of good luck that Kai said "nein"
>At the time, Kai was king of the weird interface - Bryce, KPT and Goo were all decidedly odd, leaving users with lumps of spherical rock to swivel, and glowing orbs to fiddle with just to save a simple file. Kai's interface were fun, in a Crystal Maze kind of way. He did show me one possible interface, where the desktop metaphor was adapted to have more sophisticated layers - basically, it was the standard desktop but with no filing cabinet and all your folders and documents strewn over your screen as if you'd just turned on a fan to full blast and aimed it at your neatly stacked paperwork.
The Interface of Kai Krause’s Software:
https://mprove.de/script/99/kai/index.html
>Bruce “Tog” Tognazzini writes about Kansei Engineering:
>»Since the year A.D. 618 the Japanese have been creating beautiful Zen gardens, environments of harmony designed to instill in their users a sense of serenity and peace. […] Every rock and tree is thoughtfully placed in patterns that are at once random and yet teeming with order. Rocks are not just strewn about; they are carefully arranged in odd-numbered groupings and sunk into the ground to give the illusion of age and stability. Waterfalls are not simply lined with interesting rocks; they are tuned to create just the right b...
In other words, if another person needed a logo and used the same phrase how long on average until they get a duplicate of your image?
DALL E 2 is like a low or no-code tool in that way.
The outcome may not be a "finished" product, especially as viewed by a professional designer (or web dev). However, its a heck of a lot better than a tersely written spec.
And in some cases, the product will work well enough to unblock the business, get customer feedback and generally keep things moving forward.
editing: not that it's intentional, but these things will have the same effect; way too much product even for creative works. No one will be able to make money off the product but the tools.
As someone who is incredibly terrible at graphic design but knows what they like this could be a game changer as iterations of this technology progress. I can imagine going further than images and having AI/ML generate full HTML layouts in this iterative way where you start to define your vision for a website or app even and it spits out ideas/concepts that you can "lock" parts of it you like and let it regenerate the rest.
I'm not downplaying designers role at all, I'd still go to one of them for the final design but to be able to wireframe using words/phrases and take a good idea of what I want would be amazing, especially for freelance/side-projects.
I think your art/design/craft is pretty good. Some people use pencils, some use Adobe products, you have gone out there and tried the new Dall-E medium.
Glad you thought out the usage, I am sure that when the novelty wears off that you will have that neat-as-octocat logo sorted out.
I appreciate that you appreciate the value that highly skilled designers bring to a product with their visual expertise.
However, I would like to see you A/B test the Dall E logo versus the winning designer logo. You could show odd IP addresses one logo and even addresses the other.
I think the designer would edge the robot for what you need (a logo), however, the proof is in the pudding and conversion rate.
If you lined up 100 resulting images, 99 from weekend beginners and 1 from an actual artist. I guarantee you that you would pick out the artist every time.
It might be simple to trace over an image but you are probably better getting an artist to spend 2 hours on it, it will most likely look better than 2 weeks of tracing.
People are already doing by combining DALL-E 2 with gfpgan for face restoration. So there may be a role in understanding how to combine these tools effectively.
An experienced human designer, right away, is going to ask how you want the logo to be used. That's going to have a major impact on how it's designed.
So yeah, this may be like working with a doodler, but, as the author intimated, this is far from an ideal experience in getting a professionally designed logo. This is more like "Hey, you, drawing nerd, make this thing."
Nevertheless, astonishing technology in its own right.
That's my main gripe with DALL·E as well. This missing feature makes it impossible to use for stories where the same character goes through an adventure and is present in different settings, doing different things.
Although I don't know much about how DALL·E works, I have the feeling it shouldn't be too hard to add this possibility. That would make it so much better / more useful.
No offense, but this gives me flashbacks to bad clients and non-technical managers :D
Open-ai has nonsensical censorship. Dalle might be popular right now, but open-ai won't survive if they keep up their ridiculous attempts at trying to control culture. I've already got something running on collab, new models are coming out, and midjourney just got a v3 update that blows dalle2 out of the water.
I'll have to find a different option.
Just because some asshole uses the peace symbol, does it mean the peace symbol is hateful? Anyone who claims so is dishonest
https://imgur.com/a/1IiyMJF
Maybe your perception of "logo" needs more reference points. For example, this gallery of classics in Brand Identity will be a good starting point(use the triangles on top to navigate): https://www.joefino.com/logos_html/L01_Xpand.html
There is no doubt in my mind that the next iterations of neural networks will remove all "overpaid" and "overconfident" design professionals, that's why I adapted to the reality and moved to frontend development. All of this with clear realization that everything humans can do for a production processes will be augmented and removed. The nasty "humans" always want to be paid, more and more. They want to have rights and privileges. What a hassle.:)
If someone else just needs something simple and passable there is Dalle.
And I’m sure there is every option in between where someone can use Dalle as a starting point and pass it to a pro, or a pro would even use Dalle as a way to brainstorm options.
Dalle is a tool that has empowered everyone. It shouldn’t be seen from a stereotypical luddite perspective as in your first post.
We'll hardly be able to tell the difference, if at all. Maybe it doesn't matter as long as the conversation is engaging for the human.
And what when people are certain that the machines are better in everything, who will want to chat, listen to music or watch paintings from the "lame" humans, when the robots will be the ultimate solution for every human need?
I imagine it would make it very easy to “seed” a website or a platform with initial “users” and content.
I also imagine it will be (and likely is already) being deployed to create the impression of popular support (or lack thereof) of a politician, business or policy.
I'm unsure if it's confirmation bias, but I find myself noticing weird abberations in online comments that don't seem to be ESL related. (edit: it's probably just mobile swipe typing at play)
I feel like I understand where you're coming from, but often the phrase I hear by experts (I even use this myself in my space) is, "Sure, it only took 20 minutes to do this wiring/write this code/draw this logo, but it took 5 years to know what to make." Sure, the results aren't what you'd get if you paid a professional logo designer, but if you can get close enough, it's really cutting out the X years training necessary to get to that point.
This is exactly my point. With repetition and solid design foundation comes the intuition what is the right direction towards the accomplishing of the given task.
Some will say the design is a subjective, I would argue that designers' role is to move towards objectivity and away from the idea of "personal taste".
That's why I give a link to the works of the master in this craft. This is exactly the same argument with the Copilot case. Is it capable to give some "boilerplate" solution - yes. Is this solution mediocre at best - yes.
The thing I think I like about this is I can meander through a few different concept on my own time.
https://replicate.com/laion-ai/erlich
On the bright side the result may be better, it may be easier to become and "AI usage specialist" than specializing in many different areas, the result may include many intermediate results that a specialist would find too much work to do and, with a bit a patience (like in the presented case), the task can still be done without the need of an "AI usage specialist".
Currently, I think the problem is an UI one. There should be an option to allow the user to do something like: "from the last drawing, just add this..." or "in the last drawing, change the color/size/style of this and that...". This would be probably enough to achieve what the author wanted in a much smaller number of iterations.
There is also on more thing: the costumer doesn't know exactly what he/she wants from the beginning. So, it is normal to have a few iterations until something pleasing is achieved.
Create a logo generator site, allow users to pick something very limited like industry/field from a dropdown or something, generate say 9 logos with AI generated text discriptions that fit this selection and remember which one the user picked and use that data to build a network that generates good text descriptions to feed into DALL-E 2 based on a singe item selected by the user.
This matches my view of the idea that AI will replace programmers. My value isn't in the typing and the syntax, it's in my ability to turn a spec into an internally consistent design by resolving conflicting instructions and clarifying edge cases; and sometimes in knowing what the user wants when they are unable to express it themselves.
Even if AI winds up writing all of the code, someone with the programmer mindset still needs to define the problem in a concrete manner. They'll always have a job as a "machine-talker."
We produce a lot of content and the biggest hurdle in graphic creation is the back and forth with the designer, plus the lag between writing, designing, and publishing. This would make it easy enough that the writer can include a prompt for the illustration right in the text itself.
More than the costs, I’m excited about the efficiency gains and smoother workflows.
[0]: http://dallery.gallery/wp-content/uploads/2022/07/The-DALL%C...