One must acknowledge that AI has become what it is today because of the internet. To be precise, data collected in large amounts on the internet is fundamental to the success of AI.
All AI projects essentially need large amounts of actionable data and currently, the point in time has arrived where availability of astronomical amounts of data has now coincided with significant technological breakthroughs
One example is Dall-e which can generate text to image models thanks to large amounts of data collected through the internet. It was only a matter of time until other tech giants were able to catch up with Dall-e.
Google for example has just launched Imagen which, similarly to Dall-e is a text to image diffusion model. The difference is that Imagen’s responses to written text are very photo realistic. Imagen is also venturing in text to video models. The model has not yet been fully released to the public. It is available by application and a waiting list on an app.
Even Photoshop is working on using AI. A New feature in photoshop v24 is a beta feature called backdrop neural filter. It lets you create a unique backdrop based on a description. You just type the backdrop description, select create, and pick your favorite result This new feature is still being tested and is not yet available. If they wanted to, Adobe could very well be able to create text to image models, but they simply choose not as they believe that technology should be used to enhance and benefit creators and not replace them.
