#datapoints
A lot of datapoints in the last week showing that overreliance on American tech firms is a massive security risk for Europe and the wider world
November 13, 2026 at 11:17 AM
to everyone asking me to update the model, i'm waiting for a few more datapoints to drop, but early estimates suggest beyond exponential growth
With a third Jensen Huang affair hitting today, we can now confirm recursive self improving growth in misaligned relationship capabilities

Based on current trends we could reach the affair singularity by the end of November—or even earlier
November 12, 2026 at 11:04 PM
I guess we hang with different crowds? it's good to have more datapoints!
November 12, 2026 at 6:17 PM
Fernando Mendoza I eagerly await your datapoints
The more nuanced version of the argument that Mike Macdonald regards young QBs as one of the main food groups:

QBs (Years 1-3) | QBs (Years 4+) versus the MM Seahawks:

TD rate
1.94% | 4.79%

INT rate
3.51% | 2.34%

Sack rate
9.66% | 6.3%

Success rate:
37.05% | 46.82%

ANY/A:
3.09 | 6.23
November 10, 2026 at 9:06 PM
i'm waiting for the next datapoints to drop
November 12, 2026 at 9:51 PM
based on the other decal datapoints, you can tell how hitler it is from what symbols are uses to count the family. mild hitler (ar15) to moderate hitler (crosses) to very high hitler (anime) to european extreme hitler (none because taken away by the state)
November 14, 2026 at 5:42 PM
I guess the next bottleneck is the frontend and having tools to easily select through those millions of datapoints that we can now easily UMAP 😅
UMAP 0.6 is out! A new initialization, new optimizers and faster nearest neighbour search make UMAP much faster and far better at large scale. On 3 million arXiv papers it finishes in under 5 minutes, before UMAP 0.5 has even found its nearest neighbours.
November 12, 2026 at 8:15 AM
Ars Technica did a really good article on the flaws of forensic DNA comparison. Just like you point out with fingerprints, it's just comparing a subset of datapoints. Analysis is then done by people unaware of the birthday paradox.
How weak DNA evidence railroaded—and then rescued—Amanda Knox
Not all DNA evidence is created equal.
arstechnica.com
November 13, 2026 at 7:48 AM
I’ve been wondering if my butt is getting rounder or if the mirror was lying. I dunno, the pants used to fit different lol.

Really wild when changes are noticeable across multiple datapoints. 💜
November 11, 2026 at 4:38 PM
If not absolutely no worries, I'd just love some more datapoints than the hot pot place I can't visit any more because it sounds like a symphony of dental drills since they switched to induction. :(
November 11, 2026 at 1:56 AM
Where does your data still turn back into objects on its way out?

Write-up: https://panth.vardayinitech.in/blog/1mil-datapoints-in-nodejs
Screen recordings: https://www.youtube.com/playlist?list=PLeXF8QGCGNK7MrBOweoDSd6E1L_7upSPV
November 14, 2026 at 2:19 PM
Especially, when comparing many datapoints, agents are quite fast compared to human users.
November 11, 2026 at 2:39 PM
Ouuu that would be marvelous Loo!! im also asking my friend who lives in [City that has a hot topic] if she can go in person some time to see if theyre there, but any datapoints will be helpful esp if theyre not even in stock lol.
November 12, 2026 at 9:38 PM
Forbes analyzed millions of datapoints to rank this year’s top 500 companies in the U.S. by financial performance, customer sentiment, employee satisfaction and more.
How Amazon Edged Out Apple To Become America’s Best Company
Forbes analyzed millions of datapoints to rank this year’s top 500 companies in the U.S. by financial performance, customer sentiment, employee satisfaction and more.
www.forbes.com
November 12, 2026 at 11:04 AM
More datapoints for the hypothesis that though America's long term prospects have looked increasingly hopeful since the heroic Minnesota mass protests almost a year ago, in the short term until 2028 Americans might face a lot of destabilising chaos.

With big implications for the rest of us too.
Trump calls for Democrats who have sent oversight letters and messages to Republicans to be imprisoned and orders Todd Blanche to investigate
November 7, 2026 at 8:57 PM
Consumers and the public have rights, too, and the right to build, fix and improve our own stuff is an important one. AI enables tinkering, but training AI *is* tinkering. It's taking apart trillions of published datapoints and learning from how they work to make something. new.
November 13, 2026 at 7:35 PM
Gaussian processes at scale! In 2D! Classified, systematised, optimised, and implemented in Jax! There's so much science to do now.

Very excited for this (41-page!) magnum opus from Mark Fortune, a postdoc in my team at MQ, and his open-source code at github.com/markfortune/.... Well done Mark!
New optimisations for exact, scalable 2D Gaussian process inference, with applications to astronomical time-series
Astronomical datasets often consist of a grid of datapoints in two or more dimensions, such as images, spectroscopic time-series, and imaging time-series. Gaussian processes (GPs) are a powerful tool ...
arxiv.org
November 9, 2026 at 11:18 AM
An area of significant expansion in #highered is community college baccalaureates in #teacher education, a trend that has the potential to help address the U.S. K-12 teacher shortage: buff.ly/2Tb7Psg #academicsky
DataPoints: Growing community college baccalaureates to address teacher shortage
Community colleges are taking on a growing role in preparing K-12 teachers, helping to address a shortage that disproportionately affects rural and urban communities. According to the Learning Policy…
buff.ly
November 9, 2026 at 4:32 PM
They are stkred safer than nuclear launch codes. No mentoin beyond "i think this was in the exam five years ago"

And then the exam comes along and its worded entirely differently with no datapoints given
November 9, 2026 at 11:41 PM
🔎What changes?

The revised #ESRS:
• cut the number of datapoints by over 60%
• change the materiality assessment
• add reliefs and flexibility

Double materiality and reporting on impacts, risks, opportunities and anticipated financial effects remain.
November 9, 2026 at 11:46 AM
[some-subscribed-rss] New Post: Wastebook for Friday, by Coté https://cote.io/2026/11/13/wastebook-for-friday.html
November 13, 2026 at 8:14 AM
Ok. Who is *training* their baristas? Because 2 datapoints here, several years apart.

And I hope you got caffeinated.
November 6, 2026 at 6:10 PM
This is one type how we can provide our brain more "datapoints" to reshape our responses over time. Another one is to make a hypothesi before you face something you normaly avoid.
November 8, 2026 at 4:03 PM
i know it's november and 'tis the season for substantial discounts but i'd really rather be focusing on making 3d models and grinding out datapoints which i guess the h3c will help with but still! better to be forced to weightpaint than have to deal with this poindexter baloney
November 8, 2026 at 3:11 AM