rcanand
@rcanand
Followers
297
Following
1K
Media
41
Statuses
2K
AI Engineer | LLM, vision, foundation models | local/open/private mlx/gguf/ollama vs. hosted models | Mac desktop AI | product/engg/research | #AI #LLMs
Bellevue, WA
Joined May 2008
ahai - a 100% private Mac app using local mlx AI models - found my 1541 ideas across 13447 markdown files. Thread 🧵
2
3
54
i reverse engineered 8 apps that hit $50K MRR in under 180 days and explain the pattern behind all of them 1/ find a group that spends money 2/ find a problem they repeat weekly 3/ use AI to make it work with photos/videos/inputs 4/ make the output dead-accurate 5/ replace their
96
216
2K
Behold, the bible of code foundation models. A mixed team from over a dozen institutions put together almost 200 pages of survey on different paradigms for program synthesis. 1000+ references and some nice figures. Seems like a solid resource.
6
49
535
Interesting!
The paper maps how reasoning models think, then shows their internal thinking is often shallow and messy. Reasoning models print step by step solutions, but people cannot see what thinking happens in each line. The authors define 17 mental actions, including defining the
0
0
1
What a lovely overview of agentic memory types ❤️
The hardest part of building AI agents isn't teaching them to remember. It's teaching them to forget. My colleague and amazingly talented writer @helloiamleonie just published what might be the most comprehensive breakdown of agent memory I've seen - and trust me, we all needed
0
0
2
Gemini 3 can take your visual projects to the next level. Take a look at these creations that take a look at you 👀
58
184
3K
"High schoolers excel at Oxford quantum course using pictorial mathematics" (by B Coecke, A Kissinger, S Gogioso, S Dündar-Coecke, C Puca, L Yeh, MH Waseem, EM Pothos, S Pfaendler, V Wang-Mascianica, T Cervoni, F Tomassini, V Anandraj, P Sigrist, I Khan):
arxiv.org
We are at the dawn of the second quantum revolution, where our ability to create and control individual quantum systems is poised to drive transformative advancements in basic science,...
1
11
43
I started reading papers related to context engineering and persistent memory cuz current agents suck at remembering - and RAG just doesn't work anymore. planning to share all insights from these papers as I go. these are some of the lists: - Google’s whitepaper on Context
39
66
797
This is a wonderful initiative. Been waiting for something that emphasizes nuances of openness. Thanks for doing this!
Introducing the Artificial Analysis Openness Index: a standardized and independently assessed measure of AI model openness across availability and transparency Openness is not just the ability to download model weights. It is also licensing, data and methodology - we developed a
0
0
2
Product emphasis in LLMs can (and likely will, if not already - even in open weights ones) be subtle - in pretraining, fine tuning, system prompts, tools, etc. Responses will look normal but can favor some products over organically better ones.
@koltregaskes @sama Hey, Kol. Thanks for flagging 🙏 This is not an ad (there's no financial component). It's only a suggestion to install Peloton's app. But the lack of relevancy makes it a bad/confusing experience. We're iterating on the suggestions and UX, trying to make sure they're awesome.
0
0
0
Deep eval of deep research agents :)
What's missing to build useful deep research agents? Deep research agents promise analyst-level reports through automated search and synthesis. However, current systems fall short of genuinely useful research. The question is: where exactly do they fail? This new paper
0
0
1
Love that they made them all Apache 2.0 including the large flagship. I totally see the conflict between monetizing and openness - but as a user this is better than the confusing licenses before.
Introducing the Mistral 3 family of models: Frontier intelligence at all sizes. Apache 2.0. Details in 🧵
0
0
0
0
0
0
1
0
0
I believe there is an indie AI dev approach between vibe coding and manual (or AI assisted) coding. Whatever we call it (vibe engineering?) - I share here my new mindset, and a case study after my first day of launching ahai. What is your approach as an indie AI dev?
1
0
0
How many of your ideas did you find using ahai? Out of how many markdown files? #buildinpublic #macapps
1
0
0
Pay once ($29 suggested, pay-what-you-want ≥$19) Launching today on Product Hunt + first 100 Hacker News users tomorrow get it free with SHOWHN100 (expires Thursday, Dec 4, 8 AM PT) Get it here:
rcanand.gumroad.com
Note: Responses will be slow through December 2025. Back in January.Find your ideasA completely local, private Mac app that brings together the ideas scattered across your files—so you can review t...
1
0
0