Riley Brown
@rileybrown ·
we'll see when we test it. These charts dont matter
Back
AI posts
AI posts
People and companies shaping AI. Filter your watched accounts, catch up on recent posts, or find the most liked and viewed.
Compiled by the Toolstash team · Updated
209 watched accounts · 2981 matching posts · Showing up to 3,000 recent stored posts. Counts are snapshots, not live X totals. This page refreshes every minute while visible. Last collected .
Jerry Liu@jerryjliu0 ·
Introducing the Document Parsing Protocol (DPP)
There are hundreds of different OCR models and tools, but they might all parse a document in different ways.
Some errors are very clear-cut (e.g. hallucinating text, or completely misaligning a table), but many errors are much more ambiguous (e.g. how to represent split tables sharing a header). It's an issue that plagues most parse benchmarks too, because they inherently reward a certain style even if other styles might be semantically correct.
We introduce a protocol that proposes a standard way of parsing these documents into a correct format. We've curated a list of 140 rules that cover literally every aspect of doc parsing, for instance: - tables: borderless whitespace tables, header-only tables, horizontal slicing, side-by-side mini tables - forms: key-value regions, checkboxes, checklists, ids and labels, fields outside the form, - charts: standardized chart output, multiple charts, gantt charts, line charts
You can use DPP as an evaluation rubric for LLMs when determining the quality of parsing a document.
It's not a full evaluation benchmark (yet), but an initial evaluation guide, and meant to encourage a discussion on what are the "best" ways to parse a document.
Check it out: https://www.parseprotocol.com/guide/forms-forms-and-key-value-regions
Blog: https://www.llamaindex.ai/blog/introducing-the-document-parsing-protocol
Andrew Curran@AndrewCurran_ ·
The Department of Justice is making inquiries into how the NVIDIA Hugging Face deal would affect competition.
Aaron Levie@levie ·
Agents are going to consume 1,000X more tokens than we consume today.
Because they’re going to be running in our workflows, processing all of our data, building out nearly all software, securing enterprises, and much more. To do this at scale we will need a variety of approaches at the frontier of AI.
This includes continuing to push the boundaries of closed models, while also increasing the capability of open weights models, making them more secure, as well as making them more efficient for large workloads.
Very exciting to see what @Suhail is building with Particle on this front.
Mike Taylor@hammer_mt ·
Google bought urchin tracking and renamed it google analytics and made it free. Tt was the best product on the market and completely zero cost, one line of JS to install... ...and yet mixpanel, amplitude, omniture, appsflyer etc all were billion dollar companies. People dont want to trust google to grade their own homework. Same thing with Anthropic - it doesn't matter how good claude managed agents gets, there will always be demand for cross-platform to avoid lock-in
Mia@MiaAI_lab ·
Worth revisiting
@levelsio@levelsio ·
I guess we're omarchyposting now but here's the very funky screensaver
Peter Gostev@petergostev ·
There's one model that I want that I'll never get.
Astra/Fable-based Sydney. Just imagine that.
Derya Unutmaz, MD@DeryaTR_ ·
Unbelievable new AI model Gemini 4 Carbon from @GoogleAI!
It beats Astra and Opus 5.5 on science, math, cybersecurity, and agentic coding! As I wrote before, “don’t bet against Google!”
- @levelsio
@levelsio ·
I figured out Omarchy's keybindings
Action + K
And it gives you all the shortcuts!
Daniel San@dani_avila7 ·
Trust me, connect Muse to Todoist and keep your tasks in sync
I know Muse has its own task manager, but I prefer Todoist because it’s simple, supports MCP, and lets me manage the same tasks from other AI agents like Claude Code
One task list, multiple agents
- Riley Brown
@rileybrown ·
I don’t want touch screen. I just want built in 5G.
Santiago@svpino ·
It is 2022 all over again!.
Back then, developers on this site (including me) said it was literally impossible for AI models to write working code.
Now, people know writing code is solved*, so they're moving to the next frontier: reviewing code is impossible!
tldraw@tldraw ·
We've just landed version 5.6 of the tldraw SDK.
Read the release blog for the full breakdown, or check out a few highlights below. https://tldraw.dev/blog/tldraw-sdk-5-6
- Mia
@MiaAI_lab ·
Sold at 2.5x in 1 year.
Maybe we should trade GPUs and RAM instead of stocks...
Aravind Srinivas@AravSrinivas ·
Particle is a new company focused on optimizing open-weight models to be more token-efficient and cost-efficient to serve. It's been great working together with Suhail and his team ahead of their launch! We look forward to working with them on tools like search and model orchestration.
Aidan McLaughlin@aidan_mclau ·
new gemini carbon looks great
- Mia
@MiaAI_lab ·
Google announced Gemini 4 Carbon. Pretty crazy as Argon was released only two weeks ago 😲
Per Google: - Excels at running long engineering tasks - 2x better than Argon at writing secure code. - First access goes to cyber defenders, then trusted partners, and only then to the general public.
I've only attached benchmark numbers that matter for agentic tasks and coding. Carbon leads on 5 out of 7.
I REALLY hope it actually delivers.
- Riley Brown
@rileybrown ·
Does anyone see this option in app?
Nathan Lambert@natolambert ·
I feel a little bad ranting at Gemini about the unreleased model, as going through the unofficial channels to get a frontier model out today I hear is a good bit of a mess and definitely not easy. That's why they say patience is a virtue.
- @levelsio
@levelsio ·
@dhh @omarchy Fixed the power!
Here's more about Omarchy's interface
I will be using it the next few days and weeks and tell you more what I think then
Google Gemini@GeminiApp ·
Dive into new topics with immersive explorations in Gemini.
Whether it’s tracing trade routes on the Silk Road, peering inside a jet engine, or studying the physics of lightning, each scene unfolds into interactive layers with guided explanations that let you dig even deeper.
- Aravind Srinivas
@AravSrinivas ·
Explainer videos are my new favorite way of learning things
- Santiago
@svpino ·
This is pretty cool:
You get the speed/cost/structural benefits of a decisions API like Jev.
On top of that, you also get access to current news, so you can make decisions based on the world's current state.
Check out some of the examples here:
Isomorphic Labs@IsomorphicLabs ·
Operating ML at scale is a constant battle of trade-offs: 1. Scaling thousands of models without the rigidity and complexity of ML frameworks 2. Preventing your scheduler from stalling under massive, bursty workloads We're covering both at #KubeCon NA.
- @levelsio
@levelsio ·
I opened the box form @DHH and it's the Dell XPS 14 with @omarchy loaded on it!
Set up was very easy, fast and flawless, I think it took less than a minute on first opening it
Omarchy asks you lots of questions to set it up for you, kind of like an onboarding, but not in an annoying way like Windows would do these days with too many useless questions, every question here makes sense
It sets up your Git and your AI agent of choice like Codex, Claude Code or OpenCode
Everything worked flawlessly with no bugs at all, really smooth
Then you're inside, and you see the famous Omarchy wallpaper, and then it gets a bit hard (for me at least) because I had to figure out the shortcuts
Luckily @jankeesvw helped me figure them out but I'm still learning
I'll be setting up my SSH to my VPS servers so I can work from there. And also I'd like to see if I can use Claude Code to customize Omarchy to my own liking, like generating my own apps etc.
I'll report back!
Part 2 of the video is in the reply (after I fixed the power)
Dean W. Ball@deanwball ·
A patchwork of state laws, despite having become a bit of a meme, is genuinely a risk, is a reasonably common failure mode of American federalism, and is something worth being wary of, even if that wariness doesn’t translate to anything like “blanket opposition to all state legislation.”
- Mike Taylor
@hammer_mt ·
These are all failures to have proper evals:
- they have a short half life... then you need to be able to recompile your skills for each new model - are often not necessary... then you should test against a baseline to make sure the skill is still needed - they use superficial role personas... its important to test what instructions actually give you what you want - instead focus on outcomes - that's what evals are!
If you have good evals you don't even need to look at your skills and can recompile them on your tasks every time a new model comes out (or you want to save money with a cheaper one).
Guillermo Rauch@rauchg ·
Theo cooked
Paul Couvert@itsPaulAi ·
So they've indexed 2 billion people
Not 2 millions but 2 BILLION so you can use your agent to find anyone and reach out.
Many use cases in sales, research, recruiting, etc.
People were already doing that manually... now it's way faster and more accurate.
Agentic workflows to another level.