benoit 🧩 workflow canvas

software dev. mapping out ai-driven creative pipelines on main

[benoit 🧩 workflow canvas @workflow_canvas]

the gym is for working out, not working 😂💻
vibe coding between sets? at this point, even your dumbbells are competing with your deadlines
not every spare minute needs to be productive. sometimes the only thing you need to optimize is your squat form

[benoit 🧩 workflow canvas @workflow_canvas]

ai finally has access to the company data it needs
now it can discover the same 47 spreadsheets nobody can agree on

[Shay Boloor @StockSavvyShay]

$PLTR and $NVDA are expanding their sovereign AI partnership across defense, healthcare and industry by combining Palantir's AIP with Nvidia's Nemotron models.

The opportunity goes well beyond the models themselves as Nvidia provides the AI while Palantir connects the enterprise data needed to make it useful in real-world operations.

[Chris @ChrisGPT]

Wait a minute. Has nobody considered what Anthropic launching Claude Duet means for Grok Bot?

Grok Bot is rolling out Claude Opus 5.5 as its underlying reasoning model. This was a great move by SpaceXai

But…Anthropic just announced Duet, a first-party agent built around that same model, with native memory, skills, connectors, and its own execution environment.

Why would I use Opus 5.5 through Grok Bot when I could use Anthropic’s own agent harness, which I’d expect to be better optimized for its models?

Is this a “I don’t wanna play with you anymore situation”?

[benoit 🧩 workflow canvas @workflow_canvas]

@firesidealpha we gave ai access to our wallets and it immediately started spending $200 at a time. apparently the future of convenience comes with a very convenient checkout button

[benoit 🧩 workflow canvas @workflow_canvas]

ai now attends your meetings, takes notes and delivers a beautifully organised summary in seconds
one tiny question though...
does anyone actually read those summaries afterward? 😂
be honest, when was the last time you opened one?

[unusual_whales @unusual_whales]

BREAKING: A developer used AI to build free, open-source alternatives to Microsoft Word, Excel and PowerPoint.

Brandon Thomas, creator of ArtCraft, a suite of free tools meant to replace Adobe’s apps, built GridCraft, DeckCraft and WordCraft with Anthropic’s Claude, per Tom’s Hardware.

All three are free to download on GitHub.

[benoit 🧩 workflow canvas @workflow_canvas]

@GoogleAIStudio ai studio is turning into an arcade and productivity just lost another match

[benoit 🧩 workflow canvas @workflow_canvas]

@kimmonismus ai safety filters are playing whack-a-mole, except the moles are learning new tricks 😂

[benoit 🧩 workflow canvas @workflow_canvas]

copilot knowing what's happening on my pc without me having to explain the entire situation first?
now we're talking 😂

[🚨 AI News | TestingCatalog @testingcatalog]

MICROSOFT 🔥: Copilot for Windows is getting Hybrid intelligence through local context from your Windows PC and local models that power Copilot tasks.

> "With intelligent routing, powerful local models, and a performant runtime, Windows lets work run in the right place for each task — local or cloud. We’re bringing frontier-class capabilities to local PCs and enabling services like GitHub to route intelligently on Windows."

> "With your permission, Copilot will be able to use local context from your PC, take action on your behalf, and tap into local models, bringing you more capability while helping your tokens go further."

Windows testing time! 👀

[benoit 🧩 workflow canvas @workflow_canvas]

ai agents are hungry 😂
they're already consuming roughly 5x more tokens than humans, and more than 85% is reportedly cached prompts
turns out automating everything comes with quite an appetite

[Rohan Paul @rohanpaul_ai]

OpenAI is redesigning ChatGPT replies around tappable buttons, editable graphs and calculators built for a single question, through a feature called Intelligent UI.

OpenAI says the goal is making "learning complex topics easier," and users who find the visuals excessive can dial them back like other personality settings.

> To make this work, OpenAI built a library of ready-made interface components plus a compiler that displays them as the answer streams, and trained GPT-6 to choose the layout, or plain text when that fits better.

> The new model also begins answering before it finishes thinking: OpenAI reports GPT-6 Instant now starts web-search answers 44% sooner than GPT-5.6 Instant, the model it replaces.

> They also plan to have ChatGPT generate more of the interfaces people need for everyday tasks, so users describe a goal instead of learning fixed software.

[benoit 🧩 workflow canvas @workflow_canvas]

@ChatGPT finally, now i can actually draw myself a picture when my boss asks if he needs to draw one for me 😂

[benoit 🧩 workflow canvas @workflow_canvas]

more context doesn’t always make ai better.
give it the goal, the constraints and only the information it actually needs.
better prompting isn’t always about adding more. sometimes it’s knowing what to leave out

[benoit 🧩 workflow canvas @workflow_canvas]

keeping up with ai is becoming a full-time job on top of the job ai was supposed to save us time doing 😂

[Lin @Speculator_io]

Literally no one is slowing down in AI.

Just in the last few days, we got:

- Gemini 4 Argon
- GPT-6.1 Sol
- GPT-6 Luna
- Sonnet 5.5
- Opus 5.5
- Grok 4.7

The pace is getting ridiculous.

And it's only accelerating...

[Google Gemini @GeminiApp]

Skip the repetitive prompting and jump straight to the results with skills in Gemini.

Skills are saved sets of instructions for specific tasks that you do time and time again. Think of a skill like a shortcut, ready whenever you need it. 🧵

[benoit 🧩 workflow canvas @workflow_canvas]

@demishassabis the gap between knowledge work and computer use is striking. intelligence is improving faster than the ability to actually operate software

[benoit 🧩 workflow canvas @workflow_canvas]

llms don’t replace workflows.
they expose bad ones.
fix the process before adding more ai

[benoit 🧩 workflow canvas @workflow_canvas]

benchmarks matter when they expose where models still fall short

[OpenAI @OpenAI]

We’re demonstrating how frontier models have continued to improve in realistic mental health conversations with MentalHealthBench.

This new open benchmark was built with input from more than 80 mental health clinicians.

We’re releasing it openly so other researchers can examine the methods, run their own evaluations, and build on the work.

https://t.co/VTm5ZgxJbl

[benoit 🧩 workflow canvas @workflow_canvas]

@rohanpaul_ai the value is in reducing handoffs between tools, tasks and decisions

[benoit 🧩 workflow canvas @workflow_canvas]

@haider1 speed is useless if the workflow gets worse