Last updated on 2023-04-11
I just published this post as an illustration of what is possible and happening right now with Artificial Intelligence. While I have been reading about AI and its capabilities for a few years now, today was the first time that I played with the OpenAI chat and OpenAI’s Dall-E-2 image generator. Below is a brief discussion of the fun from this morning and a quick thought on why I think curation will be critical in the future, and it is unclear to me if AI will be as good at curation as a person who is subject matter expert on the relevant topic.
First, the chat responses are relatively robust, and the style of writing is informative, if lacking in personality. But the usefulness of being able to ask narrative questions and receiving instant, coherent responses is a revolution. Google search allows you to enter an input and receive a list of links with relevant information, the AI chat can specifically answer the question. But the crazy thing is that one could produce content with this technology at an infinite scale. No longer do you need a writer to produce something passable. You can simply ask the chat AI for a 500-word article on any topic and publish it. The line between human created content and machine-made content is functionally invisible.
The ramifications of this are far reaching. For any person that writes, for any person who interacts online, for any person who searches for information. . . The future will be less like google search, and more like asking a chat bot for information and being able to trust its response.
To be fair, I asked it some legal questions and the information it provided was incorrect. But as it gains access to more data and more current information, it will be formidable. On the flip side of this will be its ability to tell what information is credible in the future, especially as information proliferates online as a result of this very technology. While it may be possible for AI to become as good at curation as a person familiar with the relevant topic, it is not there yet (based on my experience this morning).
Secondly, the image generator is interesting, but the quality of the prompt is everything. There is already at least one site (Prompt Hero) set-up to help people craft prompts that result in their desired image. The ability to craft these prompts will become standard learning for people – just as Boolean search became important to anyone dealing with searching data.
There are multiple sites that provide image prompt / creation, and I have only scratched the surface here. If you are interested in learning more, here is a list of websites, provided by OpenAI chat of course:
Can you list websites that use AI to generate images?
Here are a few websites that use artificial intelligence (AI) to generate images:
- Deep Dream: https://deepdreamgenerator.com/
- DALL-E: https://dall-e.openai.com/
- Dreamscope: https://www.dreamscope.com/
- Generative.art: https://www.generative.art/
- Artbreeder: https://www.artbreeder.com/
These websites use various AI techniques, such as deep learning and generative adversarial networks (GANs), to generate images based on user input or randomly. Some of these websites allow users to create their own images by specifying certain parameters or by using image editing tools. Others allow users to browse and explore a variety of generated images.