And of course, the Overlook hotel is actually Timberline Lodge – a place I’ve stayed and skied at over the years. I enjoyed the atmosphere so much that I attended a murder mystery party and one of their Halloween ‘Fright Night’ events.
AI generated art has caught fire. Learning how to generate the command line prompts to generate the art is still a work of trial and error. But some folks are helping you learn by giving some example prompts to help you learn what works and what doesn’t.
All the below items were 100% auto-generated and included on the page. It looks like people are exploring and sharing different prompts to generate different kinds of art.
Take an eight-minute high-speed drive through Paris in 1976 during the early hours 05:30am on a Sunday morning in August 1976 when much of Paris is on vacation. Amazing how similar, and different things look.
Using online AI art generation sites like DALLE, Midjourney, and GPT-3 aren’t free or unlimited to most folks. For example, DALL-E 2 was charging 10 cents per prompted generation attempt. Trying a few hundred prompts can quickly add up. Even using free generators like Stable Diffusion, experimenting with prompts can be time consuming.
It only makes sense we’re witnessing the rise of specialist prompt writers and online marketplaces where you can buy and sell high-quality prompts that get the desired results much faster. This saves users money on API fees and time trying to tune the prompt to get what you want.
These even have names now. A prompt engineer is a specialist adept at writing the text prompts necessary for an AI model to generate reliable outputs (such as graphics, text, or code) at a reasonable price. They can then sell the specialized prompts they generate on a prompt marketplace. These are sites where users can purchase and sell prompts. The prompt maker usually keeps 80% of the sale, and the marketplace takes a 20% cut.
Below are some of the top paid Prompt Marketplaces. Definitely worth browsing to see the amazing work that can be generated by AI art algorithms.
PromptBase– offers an amazing amount of prompts for just a few dollars:
PromptHero – seems to be geared towards higher-end generation
I think the 90’s were an amazing time for game books. This is one I’ve not seen before. It was a series of paired game books that have a first-person style play. You play against another person who is playing at the same time with the companion book. Just like in a real first-person shooter, depending on how you (and your opponent) move around, you can come up behind, in front, or to the side of your enemy as you fight against each other in the dungeon. How does that work? Questing Beast gives us a tour of how this works:
Joe Dever has allowed these books to be published on the internet and be downloaded free-of-charge. Rob Adams, Paul Bonner, Gary Chalk, Melvyn Grant, Richard Hook, Peter Andrew Jones, Cyril Julien, Peter Lyon, Peter Parr, Graham Round, and Brian Williams have also generously offered similar permission for their contributions. Project Aon hosts the books on their free website.
One of the first things you’ll run into is that you’re not going to be able to generate any images at 512×512 or larger if you have a graphics card with 8GB of VRAM or less. Even smaller if you only have 4GB of VRAM. The first/easiest method is to limit the output image size. There is also an option that splits the model into 4 parts and loads each separately (thought it will take longer), or using a more optimized/compressed set of trained model data.
So how do you do that if you have an older graphics card that only has 4GB or 8GB of VRAM? TingTingin has some tips at the end of his installation video if you are using a card with 8GB of VRAM (nVidia 3070’s for example).
Summary (at 15:45): Modify your txt2img.py and add the line ‘model.half()’ after model = instantiate_from_config(config.model) in the load_model_from_config() function.
The 1980’s Dungeon’s and Dragon’s cartoon was one of my Saturday morning favorites. I loved the adventures and magical places they went. It turns out Hasbro, like many toy companies, is starting to cater to adult nostalgia via it’s Hasbro Pulse series of toys.
Cortical Labs has taken hundreds of thousands of cortical cells from dissected rodent embryo brains, placed them in a dish, and are being taught to play Pong. The brain cells were collected from rodent embryo brains and placed in high-density multielectrode arrays. The real-time electrophysiological activity of defined motor regions was gathered to move the pong paddle. If a successful interception occurred, a predictable stimulus was delivered across all electrodes simultaneously at 100Hz for 10ms (briefly interrupting the regular sensory stimulation). If this activity did not result in an interception of the ball by the paddle, an unpredictable stimulus was delivered (150mV voltage at 5Hz for 4 seconds). This predictable/unpredictable stimulus response is apparently a natural response of brain cells.
Apparently this ‘wetware‘ hybrid approach of using real brain cells with electro-chemical interface exhibited better performance more quickly than current AI methods.
While fascinating and ground breaking I must admit it is also disturbing. It raises some serious ethical questions, especially in countries where scientific ethics differ greatly. To what degree do we want to augment our computational devices with real animal – or even human – brain cells? How could those embryonic brain cells be sourced since it seems to imply the death of the animal/person? What if we could use such devices in our own brains to enhance our own learning powers? And the bigger question: what is the end results/goals – and if we even should be considering these kinds of “wetware” brain cell augmented devices?
Reset your graphics driver – Win + CTRL + SHIFT + B
Remembering a wifi password you entered forever ago: c:\> netsh wlan show profile < shows a list of wifi profiles you have connected to > c:\> netsh wlan show profile <wifiProfileName> key=clear < shows the wifi password you last entered >
Message every single windows system on the network c:\> msg * <message>
FakeUpdate – go to the website, get on your friend’s computer, load the update screen for that OS, hit F11 to go fullscreen, and then see how long they’ll sit there before resetting. https://fakeupdate.net/