I would love to see a video on all the green and why you think they were so close but didn't.

TrainTubers: A Panel of Creators United by Transit | High Speed Rail Alliance Thom de Boo (Trains Are Awesome), Keith Davis (Car Free Ketih), and Alan Fisher join us for a virtual panel about trains, content, and more.

Coming up next! Join us on Friday, April 24 at noon CDT for a virtual panel event with some of your favorite transit creators (Trains Are Awesome, Car Free Keith, and Alan Fisher).

We'll explore the future of Amtrak, the role of influencers in train advocacy, and how to get everybody "all aboard."

2026 Inductees | Rock & Roll Hall of Fame

I've been waiting for this moment for all my life.

Phil's in!!!

rockhall.com/2026-inducte...

The drama behind an iconic landline phone YouTube video by Version History

Was just incredible fun to talk about the history of the phone system with Tim Wu on Version History (also @davidpierce.xyz, he's fine)

www.youtube.com/watch?v=Q2a1...

@edzitron.com said what they're doing is similar to burning the library of Alexandria.

Innovation. futurism.com/artificial-i...

Inappropriate and counter to the AFT's purpose.

The AFT should be zealously defending children and teachers. Instead...

#EduSky

If you follow edtech, please read this LinkedIn note from a Google Admin about Google forcing Gemini on K-12. CC @civicsoftech.bsky.social

www.linkedin.com/posts/kate-s...

Here's my Final Four in ESPN Men's Tournament Challenge

There has never been a perfect bracket. Until now. Happy #MarchMadness Day!

fantasy.espn.com/tc/sharer?ch...

When people use language, we are not engaged in the activity of simply "creating output" in the form of text. We are using it to communicate ideas, to other people.

>>

Screenshot from paper: 6.1 Coherence in the Eye of the Beholder Where traditional n-gram LMs [117] can only model relatively local dependencies, predicting each word given the preceding sequence of N words (usually 5 or fewer), the Transformer LMs capture much larger windows and can produce text that is seemingly not only fluent but also coherent even over paragraphs. For example, McGuffie and Newhouse [80] prompted GPT-3 with the text in bold in Figure 1, and it produced the rest of the text, including the Q&A format.21 This example illustrates GPT-3’s ability to produce coherent and on-topic text; the topic is connected to McGuffie and Newhouse’s study of GPT-3 in the context of extremism, discussed below. We say seemingly coherent because coherence is in fact in the eye of the beholder. Our human understanding of coherence de- rives from our ability to recognize interlocutors’ beliefs [30, 31] and intentions [23, 33] within context [32]. That is, human language use

Screenshot continued: takes place between individuals who share common ground and are mutually aware of that sharing (and its extent), who have commu- nicative intents which they use language to convey, and who model each others’ mental states as they communicate. As such, human communication relies on the interpretation of implicit meaning conveyed between individuals. The fact that human-human com- munication is a jointly constructed activity [29, 128] is most clearly true in co-situated spoken or signed communication, but we use the same facilities for producing language that is intended for au- diences not co-present with us (readers, listeners, watchers at a distance in time or space) and in interpreting such language when we encounter it. It must follow that even when we don’t know the person who generated the language we are interpreting, we build a partial model of who they are and what common ground we think they share with us, and use this in interpreting their words.

Screen shot continued Text generated by an LM is not grounded in communicative intent, any model of the world, or any model of the reader’s state of mind. It can’t have been, because the training data never in- cluded sharing thoughts with a listener, nor does the machine have the ability to do that. This can seem counter-intuitive given the increasingly fluent qualities of automatically generated text, but we have to account for the fact that our perception of natural language text, regardless of how it was generated, is mediated by our own linguistic competence and our predisposition to interpret commu- nicative acts as conveying coherent meaning and intent, whether or not they do [89, 140]. The problem is, if one side of the commu- nication does not have meaning, then the comprehension of the implicit meaning is an illusion arising from our singular human understanding of language (independent of the model).22 Contrary fn22: Controlled generation, where an LM is deployed within a larger system that guides its generation of output to certain styles or topics [e.g. 147, 151, 158], is not the same thing as communicative intent. One clear way to distinguish the two is to ask whether

Final part of screenshot to how it may seem when we observe its output, an LM is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot.

Went back to Sec 6 of Stochastic Parrots today (in the context of answering a query from a journalist) and was reminded how thoroughly we grounded that part in a discussion of language use -- y'know as communication between people.

St. John's can take out Duke, Kansas, and UConn to get to the Final Four. Go earn it!!!

Completely inappropriate. AFT and UFT should be zealously advocating for children and teachers. Not this.

Jane Hill | Substack I'm a writer, comedian and university lecturer but currently I'm mostly just dealing with ovarian cancer. I didn't know the symptoms. Now I do. And I want to make sure everyone else does, too.

I am blogging semi-regularly on Substack about what I refuse to call “my journey” with ovarian cancer.

It’s helping me and I hope it might help some other people too. substack.com/@janehill64?...

Altman is not a good faith actor. Quoting him to teachers is wasting their time.

A tech giant releasing a non-AI-coded decent device at a decent price is now a "man bites dog" story. It's as shocking as the finding out the Mavs traded Luka.

i tried editing 4K video on the $599 MacBook Neo YouTube video by patrick 2masso

"It's such an optimistic way to think about technology at a time right now where technology has felt very evil and dark."

As someone who enjoys using technology in teaching, those words, "evil and dark," are so spot on. #EduSky

youtu.be/Mo6o8RKn7jE?...

Add alt text UNEP WESR Knowledge Repository X UN Environment GPT Bein gpt-5-nano About New chat Filter by documents UN environment programme EnvironmentGPT Filter by report series Select report Ask your question 1/2008 Select audience General Public How is the thealth of sur planet coing right what does the fut What nitcators show the rent state of glot what polypr adess them? Why can circular economy he benefusal Descriptive alt text Alt text 2000 Save

At this point I am welcoming all of this stuff so everyone can just get it out of their system, stand up, shake off, and we can spend the next half-decade rebuilding normal and effective ways of creating, processing and sharing important information with people

Yikes!

A nice screen, a metal body, and a 13-inch display are miles ahead of the Chromebooks K-12 students are stuck with.

No, the new MacBook Neo isn’t a threat to Chromebooks in the classroom As it launched earlier this week, we talked about how the MacBook Neo puts massive pressure on Google’s upcoming 'Project Aluminium' devices in the consumer space. At $599, Apple has finally built a m...

This post essentially argues the MacBook Nano won't replace Chromebooks in K-12 schools because Chromebooks are junk. #EduSky #EdTech

chromeunboxed.com/no-the-new-m...

Google faces first lawsuit alleging its AI chatbot encouraged a Florida man to commit suicide Google is accused in a wrongful death lawsuit filed by the family of a man who committed suicide in October, allegedly at the direction of the tech giant's AI chatbot, Gemini.

Gemini is the latest LLM accused of generating text encouraging suicide. How many instances of LLMs generating text encouraging self-harm do we need to reconsider their use in K-12 classrooms? #EduSky www.cbsnews.com/news/jonatha...

I always come back to the fact that a *computational linguist* said this:

"Unless your use case requires a model of a distribution of word forms in text, indeed, they [LLMs] suck and aren't useful."

@emilymbender.bsky.social #EduSky

Sam Altman in Damage Control Mode as ChatGPT Users Are Mass Cancelling Subscriptions Because OpenAI Is "Training a War Machine" Sam Altman was blasted in an embarrassing Q&A while scores of ChatGPT users ditch the service in favor of Anthropic.

"Reports suggest that the DoD used Claude to select targets in Iran, meaning even Anthropic’s principled stand may be yet more theater from the AI industry."

#EduSky

futurism.com/artificial-i...

The Supreme Court doesn't care if you want to copyright your AI-generated art The highest court in the US declined to review a case about copyrighting artwork created with the help of AI.

Phew 😮‍💨

And yes, it's weird posting about my new website the day after this terrible war was launched.

Tom Mullaney Hello! I am Tom Mullaney. Teacher • 12-plus years in the classroom • he/him/his Certified by the Commonwealth of Pennsylvania to teach: Social Studies (7-12) Citizenship (7-12) Environmental Educat...

I gave my website a badly needed update. I want it to communicate that I am a teacher, and if you want someone to inject critical thinking into using edtech, you can call me for that too. www.tommullaney.com

Fingers crossed for Allam. Her defeating Foushee would be so sweet.

Hamentashen is the only apricot-flavored thing I like.

Where can I get some vegan hamentashen in the Philly area?