Peter Gasston
@petergasston.co.uk
Londoner, digital technology enthusiast, generalist, humanist, protopist, left-of-centrist, not a futurist.
Look, last post 31 minutes ago. Two goals in that time, but I can't join in the conversation. @bsky.app @support.bsky.team
September 28, 2025 at 5:38 PM
Look, last post 31 minutes ago. Two goals in that time, but I can't join in the conversation. @bsky.app @support.bsky.team
July 15, 2025 at 9:50 PM
What's the carbon footprint of using #ChatGPT?
Very small compared to most of the other stuff you do.
www.sustainabilitybynumbers.com/p/carbon-foo...
Very small compared to most of the other stuff you do.
www.sustainabilitybynumbers.com/p/carbon-foo...
May 7, 2025 at 8:16 AM
What's the carbon footprint of using #ChatGPT?
Very small compared to most of the other stuff you do.
www.sustainabilitybynumbers.com/p/carbon-foo...
Very small compared to most of the other stuff you do.
www.sustainabilitybynumbers.com/p/carbon-foo...
This is an interesting chart from it: while there’s been a lot of invention, a lot of innovation, and even a lot of adoption, we’re still in the early stages of diffusion: we haven’t yet adapted our workflows, it hasn’t become integral to the ways we work.
April 29, 2025 at 12:39 PM
This is an interesting chart from it: while there’s been a lot of invention, a lot of innovation, and even a lot of adoption, we’re still in the early stages of diffusion: we haven’t yet adapted our workflows, it hasn’t become integral to the ways we work.
April 28, 2025 at 11:18 PM
Dia is a new open text-to-dialogue model that’s taking on Eleven Labs and Google’s ‘podcast’ audio overviews. It detects emotion from the script, you can prompt it with non-verbal interjections, and has high-quality voices. Here’s a quick example I made — maybe a little manic, but very impressive.
April 23, 2025 at 11:47 AM
Dia is a new open text-to-dialogue model that’s taking on Eleven Labs and Google’s ‘podcast’ audio overviews. It detects emotion from the script, you can prompt it with non-verbal interjections, and has high-quality voices. Here’s a quick example I made — maybe a little manic, but very impressive.
Lots of games have weather effects but none has made me stop playing just to watch them as much as #AssassinsCreedShadows has.
April 14, 2025 at 10:24 AM
Lots of games have weather effects but none has made me stop playing just to watch them as much as #AssassinsCreedShadows has.
I’ve just finished reading a book about the invention, rise, and fall of the telegraph, and this excerpt feels incredibly relevant 140 years on.
![[In 1885] many telegraphers were complaining that they had been reduced to mere machines, while others decried the declining quality of those entering the profession.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:oi5cmfec5265vzoz74de5zzz/bafkreidvc6cb26bvxyynclpbzlsethifl2nmas2scatzwgra75zzntwk2y@jpeg)
April 3, 2025 at 10:10 AM
I’ve just finished reading a book about the invention, rise, and fall of the telegraph, and this excerpt feels incredibly relevant 140 years on.
At least they’re honest.
March 19, 2025 at 11:53 AM
At least they’re honest.
Please, make it stop.
March 18, 2025 at 10:41 PM
Please, make it stop.
The #GoogleGemini Flash 2.0 (Image Generation) Experimental model is pretty incredible. It generated this new image of a woman wearing a floral shirt [1] from source images of the woman [2] and the shirt [3] and a simple prompt.
March 16, 2025 at 6:12 PM
The #GoogleGemini Flash 2.0 (Image Generation) Experimental model is pretty incredible. It generated this new image of a woman wearing a floral shirt [1] from source images of the woman [2] and the shirt [3] and a simple prompt.
Testing the new Hailuo Image-01 model. Good results, although I’ve yet to throw anything tricky at it.
March 14, 2025 at 3:32 PM
Testing the new Hailuo Image-01 model. Good results, although I’ve yet to throw anything tricky at it.
You can see the difference in this screen grab comparison:
March 14, 2025 at 11:05 AM
You can see the difference in this screen grab comparison:
Google is testing a new multimodal #Gemini model that can edit images through text prompting, without regenerating a new one each time. Gemini Flash 2.0 Experimental is available to try now in AI Studio. Here’s a quick test sequence I made:
March 12, 2025 at 5:17 PM
Google is testing a new multimodal #Gemini model that can edit images through text prompting, without regenerating a new one each time. Gemini Flash 2.0 Experimental is available to try now in AI Studio. Here’s a quick test sequence I made:
RIP Skype. Mad to think that it was once practically the default verb for video calling.
February 28, 2025 at 11:44 AM
RIP Skype. Mad to think that it was once practically the default verb for video calling.
Time is a flat circle.
February 26, 2025 at 5:18 PM
Time is a flat circle.
Artificial Incel-igence.
February 26, 2025 at 10:53 AM
Artificial Incel-igence.