Tootfinder

Analysis: software and services companies have almost $30B of debt that's classed as distressed, the most in any industry apart from real estate (Reshmi Basu/Bloomberg)
https://www.bloomberg.com/news/articles/20

Tech Crunch Sees Creditors Turn ‘Violent’ to Get Repaid: Credit Weekly
The relentless rise of the Magnificent Seven makes it easy to forget that many tech firms are unprofitable and struggling with their debt burdens. Now, creditors in the space are turning against one another in a bid to get repaid.

The rampant pillaging and straight-up “appropriation” of human-generated training data from the Internet—without compensation—is a hallmark of many of today’s #AI giants. Now that they are starting to step on some wealthy toes, I think litigation is about to smack them down hard this year. NYTimes got a head start; who’s next?
“YouTube CEO warns OpenAI that training models on its videos is against …

YouTube CEO warns OpenAI that training models on its videos is against the rules
YouTube CEO Neal Mohan stated that OpenAI using its videos to train AI tool Sora would violate its terms of use.

Didn't hear about this. Interesting.
https://www.cultofmac.com/852660/apple-licenses-shutterstock-image-library-ai-model-training/

Apple licenses Shutterstock's massive image library for AI model training
To train its AI models, Apple reportedly struck a deal with Shutterstock to gain access to its library of photos and videos.

I always rush through Los Angeles Union Station - the trains and buses are so infrequent and schedule adherence so spotty that a two-minute dawdle can mean an hour’s delay.
So it is only because of a delayed train and an early-departing bus that I discovered, during the resulting 30-minute wait, that the bus plaza side of the station has two decorative fountains, a drinking fountain, washrooms, and a small monument.

I think the analogy between recovery of a sequential time code in MTL and the recovery of a sequential motor code in cerebellum is really interesting.
The cerebellum acts as the analog to the medial temporal lobe for sensorimotor memory | bioRxiv
https://www.biorxiv.org…

Good morning friends!
One more #photo from last week's hike with friends.
We were just leaving the summit when I looked back and wanted to a say 'good bye 👋' when I saw this perspective. The cross on the left and the #tree winding itself around.
I just thought this is a stron…

Sums of squares of k-term forms
Charu Goel, Bruce Reznick
https://arxiv.org/abs/2403.08697 https://arxiv.org/pdf/2403.08697

Sums of squares of k-term forms
In this paper we study the cones corresponding to sums of squares of $n$-ary $d$-ic forms with at most $k$ terms. We show that these are strictly nested as $k$ increases, leading to the usual sum of squares cone. We also discuss the duals of these cones. For $n \ge 3$, we construct indefinite irreducible $n$-ary $d$-ic forms with exactly $k$ terms for $2 \le k \le \binom{n+d-1}{n-1}$.

train_terrorists: Madrid train bombing terrorists (2004)
A network of associations among the terrorists involved in the 2004 Madrid train bombing, as reconstructed from press stories after-the-fact. Edge weights encode four levels of connection strength: friendships, ties to Al Qaeda and Osama Bin Laden, co-participants in wars, and co-participants in previous terrorist attacks.
This network has 64 nodes and 243 edges.
Tags: Social, Offline, Weighted

Didn't hear about this. Interesting.
https://www.cultofmac.com/852660/apple-licenses-shutterstock-image-library-ai-model-training/

Apple licenses Shutterstock's massive image library for AI model training
To train its AI models, Apple reportedly struck a deal with Shutterstock to gain access to its library of photos and videos.

train_terrorists: Madrid train bombing terrorists (2004)
A network of associations among the terrorists involved in the 2004 Madrid train bombing, as reconstructed from press stories after-the-fact. Edge weights encode four levels of connection strength: friendships, ties to Al Qaeda and Osama Bin Laden, co-participants in wars, and co-participants in previous terrorist attacks.
This network has 64 nodes and 243 edges.
Tags: Social, Offline, Weighted