Terretta

40 ms·

781.
by Terretta 2y ago
We, mere humans, communicate our needs poorly, and undervisualize until we see concrete results. This is the state of us. Faced with us as a client, the LLM has infinite patience at linear but marginal cost (relative to your thinking/
782.
by Terretta 2y ago
Also missing, any discussion of equity-like or profit-participatory structures from LPs (limited partnerships). Technologists joining one of these should know the "business domain" "partners" are either buying into or aw
783.
by Terretta 2y ago
> the GP Morgan head of futures department... I don't know if this is how he'd show up to his office in the bank In case you mean JP Morgan, here's the CEO, Jamie Dimon, on cover of annual shareholder letter, back in ...
784.
by Terretta 2y ago
> I now have no idea how I am supposed to dress for most things Can't go wrong with “smart casual” // not sure that helps :)
785.
by Terretta 2y ago
I noticed the term "syllable" doesn't appear in the paper. If you use single utterances to convey things, and mostly are uttering across treetops far away to share things others aren't seeing directly (much less chatty
786.
by Terretta 2y ago
Abstract Compositionality, the capacity to combine meaningful elements into larger meaningful structures, is a hallmark of human language. Compositionality can be trivial (the combination’s meaning is the sum of the meaning of its parts)
787.
In the Calls of Bonobos, Scientists Hear Hints of Language(nytimes.com)
2 points by Terretta 2y ago|2 comments
788.
by Terretta 2y ago
What is "Custom prompt to let us know what you're reading"?
789.
by Terretta 2y ago
> If you are going to use human brain cells to memorize protected content and sell it as a product, that's still an issue based on current copyright laws. And yet, that's all most billable hours at McKinsey, BCG, KPMG, are
790.
by Terretta 2y ago
> violate copyright when training If we could incrementally update our own brains by swapping cells for chips, what percentage of our brain has to be chips before us learning from a book is a violation of copyright? When learning to re
791.
by Terretta 2y ago
Speaking as someone who enjoys reading after dark, I like lanterns, but I consider the ecosystem dire. No one seems able to propose a solution to the problem of how do I keep this lantern lit all night without soot and fuel on my hands. At
792.
by Terretta 2y ago
NYT among others found ads placed with content outperformed ads attempting to target users. Also avoided awkwardly advertising things that shouldn't be advertised alongside certain content.
793.
by Terretta 2y ago
There's a difference between writing in your journal and doing research. One of these is just "reporting the news". Even the magazine is not called "Consumer Research".
794.
by Terretta 2y ago
"They continue to fabricate links, references, and quotes, like they did from day one." - "I ask them to give me a source for an alleged quote, I click on the link, it returns a 404 error." Why have these companies no
795.
by Terretta 2y ago
There are (were*) armies of low-paid humans informing RLHF in response to awful inputs. https://slate.com/technology/2023/05/openai-chatgpt-training... And aside from that, we know there is more content in ob
796.
by Terretta 2y ago
And why not contribute to Aider.chat at https://github.com/Aider-AI/aider
797.
by Terretta 2y ago
Perhaps the author meant "Apple is succeeding too hard". :-)
798.
by Terretta 2y ago
See also: https://simonwillison.net/2024/Oct/25/pelicans-on-a-bicycle/
799.
by Terretta 2y ago
> Because few ever write diagrams in SVG. SVG is an output format, not an input format... Aside from "try Inkscape", that sounds like a human problem not an LLM problem. LLMs output what they input, and if diagrams in blog ar
800.
by Terretta 2y ago
This is a novella that just keeps getting twistier, going deeper, and turning out more rewarding. And clearly all these "reasoning" LLMs trained heavily on Eric Draken's refreshingly introspective monologue full of interjecti
801.
by Terretta 2y ago
Craft has collections for tables of data, also has a variety of other data modeled things like TODOs: https://www.craft.do/
802.
by Terretta 2y ago
If you're familiar with the space, ever heard anything that ticks all these boxes: (a) git versioned (b) CRDTs (c) WYSIWYG but Markdown (esp GFM) (d) front matter + markdown able to be SSGd (e) comfortable UI for not-devs (f) native (S
803.
by Terretta 2y ago
That said, calling the video conferencing capability "Visio" is à la banane.
804.
by Terretta 2y ago
From the article: You can order GLP-1 peptides from China for cheap. Once you have the peptide, all you have to do is put it in the right amount of bacteriostatic water. In theory this is no harder than any other mix-powder-with-water task
805.
by Terretta 2y ago
Such a shame products like this hide features behind Call Us or Let's Chat buttons. Just let people buy it.
806.
by Terretta 2y ago
Due to pricing model, due to cost CloudCraft is effectively unusable for small shops using accounts as an information boundary (per best practices for certain sectors). Doesn't mean small firms wouldn't want to. Just, it's n
807.
by Terretta 2y ago
On the contrary, even telling people which billboards are posted about what, and how to get to them to look at them, is "how it works". But the courts will get to clarify (in today's news): https://www.reuters.com&
808.
by Terretta 2y ago
> the companies training the models are stealing the content from the internet Are you stealing a billboard when you see and remember it? The notion that consuming the web is "stealing" needs to stop.
809.
by Terretta 2y ago
> when it comes to real people, they get sued into oblivion for downloading copyrighted content, even for the purpose of learning. Really? Or do they get sued for sharing as in republishing without transformation? Arguably a URL prov
810.
by Terretta 2y ago
Is the book online and accessible to your eyeballs through your open standards client tool, such that you can learn from seeing it?

More ›