Tootfinder

Opt-in global Mastodon full text search. Join the index!

No exact results. Similar results found.
@jonippolito@digipres.club
2025-11-17 14:46:31

Every post you write is first read by a machine. Should the nature of writing change to accommodate that? I look at the pros and cons of adopting an AI-friendly grammar linkedin.com/posts/jonippolito

A female author typing on a computer keyboard with a futuristic machine version of her on the computer screen looking back at her, created with Leonardo.ai.

#writing #journalism #aiineducation #ailiteracy #llm | Jon Ippolito
Who are we writing for in the age of AI? Good writers write with readers in mind, and most assume those readers are human. As much as we might hope this is the case, however, the reality of the Internet age is that the first readers of our texts are machines. This didn’t start with large language models. Back in 2009, Facebook replaced its chronological feed with an algorithmic one, and now “For You” screens are the default in most social media platforms. Documentaries like The Social D…

@v_i_o_l_a@openbiblio.social
2025-10-16 11:52:31

bei "Fragen Sie Dr. Bopp!" gabs jetzt "was mit bibliothek": blog.leo.org/2025/10/11/wenn-b 🙃

@mszll@datasci.social
2025-10-03 09:49:37

🤩 Fantastic new network plotting package available in Python by Fabio Zanini. The package supports both #networkx :networkx: and #igraph :igraph: networks, and has a wide variety of styling options.

@hex@kolektiva.social
2025-11-06 13:32:50

Oh hey, more folks pointing out that the shutdown may just be the goal #USPol

@arXiv_csCV_bot@mastoxiv.page
2025-10-09 10:01:11

Adaptive Stain Normalization for Cross-Domain Medical Histology
Tianyue Xu, Yanlin Wu, Abhai K. Tripathi, Matthew M. Ippolito, Benjamin D. Haeffele
arxiv.org/abs/2510.06592

Adaptive Stain Normalization for Cross-Domain Medical Histology
Deep learning advances have revolutionized automated digital pathology analysis. However, differences in staining protocols and imaging conditions can introduce significant color variability. In deep learning, such color inconsistency often reduces performance when deploying models on data acquired under different conditions from the training data, a challenge known as domain shift. Many existing methods attempt to address this problem via color normalization but suffer from several notable dra…

@aral@mastodon.ar.al
2025-10-24 06:54:42

If you read one thing today…
1. You should read more
2. Make it this 👇 mastodon.social/@danielquinn/1

Daniel Quinn (@danielquinn@mastodon.social)
David Suzuki, one of Canada's greatest voices on climate and the biosphere says that it's too late and that the fight against climate change is lost: https://www.ipolitics.ca/2025/07/02/its-too-late-david-suzuki-says-the-fight-against-climate-change-is-lost/ "The units of survival are going to be local communities, so I’m urging local communities to get together." As someone who has loved Suzuki for a long time, this interview was both brutal and refreshing. He's basically saying that we'…

@jonippolito@digipres.club
2025-12-09 14:11:46

We've updated the What Uses More app to reflect last week's finding by Luccioni and Gamazaychikov that "reasoning" mode increases energy and water usage by 30x. The study casts doubt on the improved efficiency AI companies are claiming for newer models

A screenshot from the What Uses More app, showing a chart with 30x more energy usage for reasoning models.

#ailiteracy #aiethics #ecology #environment #climate #sustainability | Jon Ippolito
We’ve updated the What Uses More app to reflect last week’s finding by Dr. Sasha Luccioni 🦋 and Boris Gamazaychikov that “reasoning” increases AI energy and water usage by 30x. The AI Energy Score project has released version 2 of its public leaderboard, now including a benchmark for models that perform more complex tasks via “reasoning” modes that involve intermediate steps, such as chain-of-thought. The results are striking: AI models with reasoning on average consume 30x mor…

@radioeinsmusicbot@mastodonapp.uk
2025-10-04 13:53:38

🇺🇦 Auf radioeins läuft...
Jonathan Jeremiah:
🎵 Kolkata Bear
#NowPlaying #JonathanJeremiah
open.spotify.com/track/6GVyNJO

@jonippolito@digipres.club
2025-10-09 12:49:05

The surge in watermark removers within days of Sora 2’s release reminds us that most AI detection is just security theater at this point.
linkedin.com/posts/jonippolito

Two images of an Asian man with a hat, one with a Sora watermark and the other with the Sora watermark removed.

#sora #aiethics #aivideo #ailiteracy #deepfakes #openai #cybersecurity | Jon Ippolito
The surge in watermark removers within days of Sora 2’s release reminds us that most AI detection is just security theater at this point. Detection advocates will counter that sure, visible marks like the little Sora “cloud” can be cropped or Photoshopped, but embedded watermarks like Google’s SynthID are harder to rub out. Unfortunately even steganographic watermarks can be scrubbed by screenshotting, model-to-model laundering, or just serious editing. An imbalance of incentives mean…

@jonippolito@digipres.club
2025-11-01 12:33:05

Google has teased 60-second AI videos from Veo, which sounds impressive until you realize that requires not 10x, but 100x the energy of a 6-second clip.
Analysis confirming quadratic scaling of AI video footprint: linkedin.com/posts/jonipp…

News headline from Techradar about rumored minute-long AI videos coming from Veo 3.1

New research reveals shocking energy use of AI video | Jon Ippolito posted on the topic | LinkedIn
I've updated the What Uses More calculator with new research this week that makes the environmental footprint of AI video potentially far higher than I thought. On Tuesday Julien Delavande, Régis Pierrard, and Dr. Sasha Luccioni published important research showing that doubling the duration of a video generated from a text prompt doesn't just double the energy required, but in fact quadruples it. So if a 3-second clip uses 130 watt-hours (the equivalent of running a light bulb for two hou…