Ben Cobley

43 posts

Ben Cobley

Ben Cobley

@cobley_ben

Creative technologist @GoogleDeepmind | views are my own

Katılım Eylül 2024
117 Takip Edilen145 Takipçiler
Ben Cobley
Ben Cobley@cobley_ben·
@matvelloso 💯 we'll be telling our grandkids we used to get http from a server
English
0
0
1
44
Alp
Alp@alpniks·
@GoogleDeepMind The speed of this is crazy, it seems to be able to replicate a bunch of static sites I would create for my friends instantly. I'm also a little surprised by the idea of streaming HTML code working this well without any perceived latency. Props to @cobley_ben
English
1
0
3
1K
Google DeepMind
Google DeepMind@GoogleDeepMind·
Watch how fast Gemini 3.1 Flash-Lite can generate websites. ⚡ This browser creates each page in real-time as you click, search, and navigate. Give it a try → goo.gle/4t9In1R
English
149
364
3.2K
834.4K
Ben Cobley retweetledi
Jeff Dean
Jeff Dean@JeffDean·
Here's a side-by-side speed and accuracy comparison comparing Gemini 3.1 Flash Lite against our older Gemini 2.5 Flash model. Not only is the new model significantly faster in terms of tokens/s, it also needs many fewer tokens to accomplish complex tasks (~1/3 as many in this case).
English
3
11
80
13K
Ben Cobley
Ben Cobley@cobley_ben·
first output that made us realise Gemini 3.1 is surprisingly good at animation Prompt: "Generate an SVG of a seal bouncing a beach ball on its nose, animated"
GIF
Ben Cobley@cobley_ben

I made this video after discovering that 3.1 Pro is crazy at animation Awesome to get a shoutout from the original pelican artist, thank you @simonw! Always trying to highlight new capabilities we've found + show them side by side (see prev tweets) Your benchmark has been a great source of inspiration for us, thank you!

English
0
2
7
513
Ben Cobley
Ben Cobley@cobley_ben·
I made this video after discovering that 3.1 Pro is crazy at animation Awesome to get a shoutout from the original pelican artist, thank you @simonw! Always trying to highlight new capabilities we've found + show them side by side (see prev tweets) Your benchmark has been a great source of inspiration for us, thank you!
Ben Cobley tweet media
Jeff Dean@JeffDean

Today, we’re continuing to push the boundaries of AI with our release of Gemini 3.1 Pro. This updated model scores 77.1% on ARC-AGI-2, more than double the reasoning performance of its predecessor, Gemini 3 Pro. Check out the visible improvement in this side-by-side comparison, showing Gemini 3.1 Pro’s crisp animation built with pure code. Read more about today’s 3.1 Pro update: blog.google/innovation-and…

English
0
3
6
806
Ben Cobley
Ben Cobley@cobley_ben·
not just pelicans, Gemini 3.1 aces SVG animations for your website too
English
0
1
10
772
Ben Cobley retweetledi
Google AI
Google AI@GoogleAI·
Gemini 3 Deep Think is getting an upgrade 🧠 By blending deep scientific knowledge with advanced engineering utility, Deep Think now moves beyond abstract theory to drive practical applications. Researchers are already using it to accelerate their work in the real world: — Materials science: A university lab used Deep Think to optimize the growth of complex crystals that are candidates for high-temperature semiconductors — Mechanical engineering: A researcher demonstrated how Deep Think can be used to iterate on physical prototypes at the speed of software. When applied to things like assistive devices, this pace means faster improvements for life-changing technology (more information on this use case in the video below!) The updated Deep Think is available in @GeminiApp for Google AI Ultra subscribers.
English
68
127
877
194.5K
Ben Cobley retweetledi
fofr
fofr@fofrAI·
I asked for a 3D spinning 'G' Google logo. Gemini 3 Flash was 3x faster, used 50% fewer tokens and gave much better results. Gemini 3 Flash: 6.6k tokens, 38s Gemini 2.5 Pro: 13k tokens, 108s You can use this AI Studio app to do your own comparisons: aistudio.google.com/apps/bundled/a…
Logan Kilpatrick@OfficialLoganK

Introducing Gemini 3 Flash, our frontier intelligence model, available at scale for everyone. It excels at coding, tool calling, and is stronger than 2.5 Pro across most metrics!! ⚡️ Available in the API at $0.50 in / 1M tokens and $3.00 out / 1M tokens across.

English
6
17
294
47.2K