Google just dropped a LOT. Gemini 3.1 Pro is here with a 77.1% ARC-AGI score (that’s above the average human baseline), Lyria 3 lets you generate music from images, and a new feature called Photo Shoot creates AI product photography. On the video side, Magnific finally drops creative video upscaling — I test it on Seedance and CG footage with some wild results. And Midjourney is gearing up for V8 with early test images, a new editor, and a bigger V2 video model coming later this year.
🔑 In This Video:
— Gemini 3.1 Pro benchmarks, availability, and hands-on coding test
— Lyria 3 music generation with multimodal image-to-music
— Google Photo Shoot for product photography
— Magnific creative video upscaling: what works, what gets weird
— Midjourney V8 early look, new editor, and V2 video model news
— Possible Veo 4 timing hint
📌 All tools tested in this video are available now or launching within the next week.
—
—
#Gemini31 #Google #AIVideo #Magnific #Midjourney #MidjourneyV8 #Lyria3 #AIMusic #AITools #GenerativeAI #TheoreticallyMedia
Chapters
0:00 — Intro: A Seedance-Free Day
0:46 — Gemini 3.1 Pro: Benchmarks & ARC-AGI Score
1:53 — Humanity’s Last Exam Results
2:30 — Gemini 3.1: Where to Access It
2:53 — Hands-On: Missile Command Clone Test
3:48 — Lyria 3: Google’s AI Music Generator
5:04 — Lyria 3: Image-to-Music Test
6:31 — More Lyria Thoughts & What’s Coming
6:41 — Google Photo Shoot: AI Product Photography
7:55 — Magnific Creative Video Upscaling
9:30 — Magnific: Controls & Settings Breakdown
9:52 — Testing Magnific on Seedance Outputs
10:47 — Magnific Quirks: Hallucinated Faces & Artifacts
11:22 — Will Smith vs Spaghetti Monster Upscale
11:56 — CG & Game Footage Upscaling (Witcher 4)
12:46 — Magnific on Freepik + World Labs Footage
13:54 — Midjourney V8: Early Look & Text Rendering
14:55 — V8 New Editor, Interface & V2 Video Model
15:44 — Midjourney V8 Launch Timing
15:55 — Outro & What’s Coming Next Week
source
Leave a Reply