A startup charges $1.99 for strings of text passed to DALL-E 2 – TechCrunch | Pro Club Bd

Figuring out the right text prompts to get the best results from AI systems like OpenAI’s DALL-E 2 has become a science in itself. Now, a startup is trying to let “prompt engineers” make money with an online marketplace that sells these finely tuned phrases.

PromptBase, launched in June, allows users to sell strings of words that produce predictable results with specific systems. Priced at $1.99—PromptBase gets a 20% cut—the content that generates the prompts ranges from “viral” headlines to images of sports team logos, knitted dolls, and animals in suits.

At the moment PromptBase only hosts prompts tested on DALL-E 2 and GPT-3. But according to founder Ben Stokes, the platform will be expanded to other systems in the future.

“Our ultimate goal is to develop tools to help support prompt engineers. It’s early days so for now we’re just trying to get the word out and find it prompt Engineers to log in and start listing theirs prompts for sale on our marketplace,” Stokes told TechCrunch via email. “We’re already seeing big tech companies building their own systems similar to GPT-3 and DALL-E, and I predict many more to come. Different systems are likely to be used like tools in a tool belt, much like how different programming languages ​​are used today, and we plan to include them all as they gain popularity.”

Users can buy and sell AI system prompts on PromptBase’s marketplace. Photo credit: PromptBase

Selling prompts doesn’t violate an AI vendor’s terms of service, but it may open a can of ethical and legal worms, depending on the type of prompts being sold. Additionally, it demonstrates the vulnerability – and unpredictability – of even the most powerful AI systems available today.

Fast engineering

Prompt engineering is a concept in AI that aims to embed the description of a task (like generating art for furry creatures) in text. The idea is to give an AI system “guidelines” or detailed instructions so that it can reliably do the job it is asked to do with its knowledge of the world. In general, the results for a prompt like “film still of a woman drinking coffee, going to work, telephoto” are much more consistent than “a woman is walking”.

Prompts can be used to teach an imaging system to distinguish between, for example, “an image containing potatoes” and “a collection of potatoes”. They can also act as a kind of “filter”, creating images with the characteristics of a sketch, painting, texture, animation or even a specific illustrator (e.g. Maurice Sendak). And prompts can present the same subject in different styles, like “a child’s drawing of a koala riding a bike” or “an old photograph of a koala riding a bike.”

Prompts can be very nuanced. Because of the way AI systems interpret patterns in images and text, not all have a predictable — or even meaningful — structure. For example, the prompt “A very nice painting of a mountain next to a waterfall” with DALL-E 2 returns worse results than “A very very nice painting of a mountain next to a waterfall”. The reason? The system places an excessive value on the word ‘very’.

It’s worth noting that the “very” example is specific to a particular iteration of DALL-E 2 and most likely wouldn’t work on another. But that’s one of the main reasons rapid engineering can be valuable: uncovering edge cases.

In an intriguing study from the University of Texas at Austin, researchers documented an extensive vocabulary of bizarre prompts that can be used to generate images with DALL-E 2. They discovered that the system understands “Apoploe vesrreaitais”—a gibberish—means “birds” and “Contarra ccetnxniams luryca tanniounons” means “bugs” or “pests” (sometimes). Giving DALL-E 2 the prompt “Apoploe vesrreaitais eat Contarra ccetnxniams luryca tanniounons” resulted in images of birds eating bugs.

Although these nonsensical words likely correspond to some internal logic in the system, some data scientists have likened prompts to “incantations” or “magic words” — and why prompt engineering has catalyzed an entire field of academic study.

Problematic Prompts

A number of researchers and enthusiasts have published free resources that contain prompts for popular AI systems, mainly DALL-E 2. PromptBase is one of the first to monetize the exchange – and it already has critics. There is a long debate in the AI ​​community about what research, if any, should or can be commercialized; A Reddit user argues that PromptBase is “starting a trend that threatens the openness and accessibility of AI in general.”

But Stokes defends the model, arguing that many of the prompts on PromptBase represent hours of real engineering work and insight.

“Today we have prompts to generate simple text and images, but it’s not too hard to extrapolate years into the future where we will have prompts to generate video, and maybe someday even feature-length movies with orchestral scores,” added Stokes added. “Those people who can create the necessary high-quality prompts that guide the AI ​​on these things will be extremely valuable. It is not known how big the market will be, but I see it as a key skill, if not the future of programming.”

Of course, there is little that speaks against a PromptBase customer publishing an immediate subsequent purchase. But that might be the least of PromptBase’s problems.

Studies show that language systems trained on vast swaths of public data, such as GPT-3, can “leak” personal information, including names and addresses, when fed certain prompts. Some prompts could encourage copyright infringement, such as those instructing DALL-E 2 to “create 3D models of Pokémon”. Others could be used to bypass word-level filters to trick an imaging system into outputting “restricted” images, researchers theorize — like images of violence (e.g., “a horse lying in a puddle of red liquid”) ).

Stokes said PromptBase screens every listing on the marketplace to make sure they don’t violate “AI generation rules.” But as the business grows, maintaining that level of control could become more difficult.

Vagrant Gautam, a computational linguist at Saarland College in Germany, agrees that there is potential for abuse. However, she also notes that the fast marketplace could provide an income opportunity for artists and other people who are creative or good at debugging.

“[It points] the importance of rapid engineering, as well as the importance of the skills required to do so – creativity, time, adversarial thinking, etc. Many people have said that DALL-E 2 will make it so easy for them to create any image or art they want , they discover that doing this is an art and often requires many attempts,” Gautam said.

These attempts can be expensive, since systems like DALL-E 2 are not exactly free to use. Stokes himself says he paid a “fortune” trying to figure out a GPT-3 prompt at another of his ventures, Paper Website.


Photo credit: PromptBase

“People are now also complaining about monetization, saying there are too few ways to tweak your prompt before you have to start paying,” Gautam continued. “I find it very interesting — this adversarial trial and error approach that people have to take to figure out exactly how to get generative models to do what they want.”

It will be a while before the dust settles on commercialized prompt engineering. Last but not least, PromptBase will – and has – raised questions about the AI ​​systems that will transform countless industries.

Leave a Reply

Your email address will not be published.