TODAY'S AI NEWS: Ideogram just dropped a new SOTA image generation model! Plus, more news from Alibaba, Microsoft, OpenAI, xAI, and Amazon. Here's everything you need to know:
MULTIMODAL AI
-
ChatGPT 4o posts screenshots of its own outputs
By
–
ChatGPT 4o generates a screenshot of it posting the screenshot it’s generating to X dot com, where it is poorly received
-
Generate AI Characters in Multiple Art Styles and Formats
By
–
Make me a South Park character
Make me a 3D voxel character
Make me a pixel art character
Make me a Minecraft character -
Vision AI Inspects Energy Infrastructure for Utilities
By
–
@BuzzSolutions1
, a member of the #NVIDIAInception program, is using vision #AI to help #utilities inspect, maintain, and secure critical energy infrastructure, including: Transmission lines Distribution poles Substations Listen in: https://
nvda.ws/4hMVrUw -
Ideogram and Reve excel at image generation while 4o dominates restyling
By
–
I actually think Ideogram and Reve are better at the initial generation. I'm loving 4o's ability to restyle and reimagine.
-
Converting sketches to logos using GPT-4o
By
–
You can turn any sketch into a logo with 4o (transparent PNG):
-
Insert any product inside any environment with 4o image generator
By
–
Insert any product inside any environment with 4o image generator:
-
ChatGPT Successfully Generates Accurate Pet Picture from Description
By
–
Nailed it. I told ChatGPT to make a picture of my pet. It asked me a handful of questions and came up with this. It got the breed mix, the ears, and the fluff in the tail. Only thing wrong is the tail is too long.
