Bill Gates on AI, the Pope Drip, Adobe gets involved
Generative AI and particularly Image generation AI is all the rage these days and things are about to get even more interesting. Also, Bill Gates has some words for you….
Last week Bill Gates wrote about AI, and the world was finally at peace, the Age of AI has been allowed to begin.
Of course, he has been meeting with the OpenAI team since 2016, so nothing is a surprise to him. In his blog post about AI, he stays away from any controversial remarks, but he does make some good points on the changes Generative AI brings.
What I like the most about the text was the advice to “keep in mind that we’re only at the beginning of what AI can accomplish.” and whatever limitations it has today will be gone before we know it. I think that is a very important aspect to not be missed in the whirlwind of amazing discoveries that is AI these days.
He also makes the case for looking at AI as something that can improve people’s lives rather than end them with the added note that governments should get involved ASAP.
Also, his remark about Education needing to adapt to new tools like ChatGPT is 100% valid. Education adapted to the internet, Wikipedia, and other sources of information and content and it will again.
You can read his full blog post here.
In other news, Generative Ai for photos is getting crazy! Midjourney V5 is so good that some very hard-to-spot fakes made rounds all over the internet. Besides Trump getting arrested or Elon Musk holding hands with AOC, I think the most viral Midjourney creation was the Pope wearing a trendy winter jacket.
That Pope photo really is something else, outrageous, yet still believable. Be very careful about what you see on the internet starting now, things are about to get very messy
And while Midjourney v5 was blowing people’s minds, Microsoft was trying to get a piece of the attention pie with Bing Image Creator, a text-to-image generation service based on Dall-E.
I got to play a little bit with the new service and it’s quite good, to be honest. They’re catching up to Midjourney rather quickly. You can find below some examples from the prompt “a cold bottle of beer in the sun, on a table, mountains setting, people laughing”.
You can play with Bing Creator it at https://www.bing.com/create
Also last week Adobe joined the Generative AI party with Firefly, a(nother) text-to-image service that will eventually be available via APIs on various platforms to enable customers to integrate into custom workflows and automation. You can sign up for the beta here.
What I found particularly interesting among all the information in Adobe’s press release for Firefley’s launch, was the introduction of the “Do Not Train” tag for creators who do not want their content used in model training. I think it’s the first use of this term by a well-established company involved in content creation. Nice!
And if you thought that the image generation novelties are over, there is one more launch I want to mention, Nvidia Canvas. The kings of GPUs have launched a desktop app for artists.
The app allows you to turn simple brushstrokes into realistic landscape images using AI. It’s not as spectacular as text-to-image, but I guess it does have its uses and it also integrates with Adobe Photoshop. You can download the beta (for Windows) here.
PS: Of course, there are more Generative AI tools being pushed to production every day, but most of them still need work before they get featured here. Thinking of doing a showdown between the main competitors for image generation soon.