I've been asked to write an article about AI for a construction magazine. the final version has been shortened to fit on a page, here's the link:
https://online.fliphtml5.com/orwf/iaoa/#p=1
and here's my full article:
https://docs.google.com/document/d/1V7iDF19IDf9CV2miq1dt4CJVESRUTB3b_U7dAL-KcCo/edit?usp=sharing

Google a few days ago released a new video model Veo3 and a service called Flow, where you can create movies and clips. It's not available in Europe yet, so I haven't tried it myself (although it's probably possible via VPN), but other people's videos look awesome. In addition to the quality of the video itself, you can generate sound and character speech, including dialogs. Here is a short tutorial on how to use it:
https://www.youtube.com/watch?v=9nVEfjmDlVk&ab_channel=Google
#ai

If you don't know what to do in the next couple of years, here's the thing, Design Morphine has uploaded their webinars on Youtube (the one we did with Stepan Kukharsky is also there):
https://www.youtube.com/playlist?list=PLIG8fMLkHigepUf9ycIJHiLOGydY43rKT
#education

At the #comfyui workshop, which ended back in March, we had a course project with a deadline of June 1, and now participants are slowly submitting their work. Part of this assignment could be training your own models. Ekaterina Zakablutskaya chose brick buildings as her topic; here is the model she trained for this task:
https://civitai.com/models/1615218?modelVersionId=1827992
#education

Today new models from Black Forest lab were released, the general name is FLUX.1 Kontext. the essence of work there is like Gemini and ChatGPT, that is, these models can understand how to modify the image simply by text instruction. in #comfyui all this is available through API, so it will not run from your computer. description here:
https://blog.comfy.org/p/flux1-kontext-api-node-in-day-1-workflow