Hacker Newsnew | past | comments | ask | show | jobs | submit | tzm's commentslogin

People want local AI, but only if UX is good. Tooling/harness quality may matter as much as model quality.

I think the future will probably be a hybrid of:

1. local AI for simple, private, everyday tasks

2. online AI for very hard or long tasks


The Clippy app someone made and posted here a while back is the perfect average person LLM interface;

https://felixrieseberg.github.io/clippy/


This is so good. Wiring in small models for a variety of tasks would make this absolutely sing.


it's a self enforcing loop

local LLMs builds tool that does exactly what user wants, how it wants it, which is bext UX

this becomes AI literacy

LLMs already nicely bridge the gap form "I want this" to "here's a local page that does it".

examples of tools i have built that requires almost very low tech knowledge * push a button on my phone to take screenshot in my mac (when i watch videos) * help me exercise, gamify it for me * "help me track time spent online to how it impacts what i do in real life, built a tool that rewards and me points me towads things that make me DO things online" * i want to improve my writing, give me exercises and build addiitonal tools (leading to an "append only" digital keyboard i use to exercise )

local AI can already create these tools, and no external company is ever going to beat me/the-user because instead of getting features i don't want, or that almost do what i want, or that do something that advantages the company they just do what I want

Repositories of tools-as-ideas created by others are quite often just index.html and ... that's all? manage data in localstorage, end of it?

Online inferences is still needed for large data (audio/video/images) processing. For now? we don't know, history suggests we'll have the capabilities to do that locally "soon". Or maybe not :)

The main issue is "online for collaboration". Not same user across different devices, that is easy. MeteorJS-style approaches (making local copies of part of dbs, reconcile to remote/origin) seems to be an interesting possibility at small scale, since once you have the right primitives in place you can go horizontally everywhere.


The UI is already great.

I can’t wait to run my models locally. The sooner I can do my shit without some American mega corp gulping down all my data, the better.


I fear that easier it gets to run models locally, more expensive all the hardware gets. So at the same time it gets further and further. You should have bought the hardware yesterday.


The more expensive it gets, the higher the incentive for more competition in the hardware space.


The thing is that it is something which takes so long time. E.g. why Taiwan is still so important?


Does it improve things or break things?


This is a misleading post, of course they can/will. Elon's comment is referencing a specific a memory bandwidth limitation of HW3 (1/8 memory). The solution is to upgrade to > HW4 as a retrofit.


It's a new computer and new cameras. What's your plan for doing the upgrades? Tesla says it's too hard and they still don't have a plan: https://electrek.co/2026/04/22/tesla-will-build-factories-ju...

And HW4 is already obsolete: https://electrek.co/2026/04/23/tesla-hw4-plus-upgrade-will-h...


Disney seems played out. Still feels early for YouTube.


I think zooming is effective when it's used in isolation for discrete things. It does add a sense of delight, but there is a functional usefulness of this that I'm trying to wrap my head around.. perhaps a transition effect for an immersive demo, etc.. nice work.


Thanks! Yeah, immersive demos and dashboards are where it seems to click best.


Copilot Health | https://copilothealth.com | REMOTE (SF preferred)

We’re building the operating system for medical AI agents—enabling at-home diagnostics with computer vision and voice AI to deliver continuous, equitable care in a broken system. Targeting asynchronous care, we serve hospitals, rural and specialty clinics with >85% diagnostic accuracy and >70% patient engagement.

Mission: fix healthcare access and equity.

Tech: PyTorch, 3D reconstruction, SAM, React, Swift/SwiftUI (iOS), Kotlin (Android), React/Next.js (web), GCP, FHIR, Temporal, MemGPT. HIPAA-compliant (SOC 2 in process)

Roles

  - AI Platform Engineer: Scale vision-language models, voice AI (emotion/intent), 3D/hyperspectral pipelines, multi-agent systems, FHIR integrations.
  - ML Engineer: Fine-tune VLMs (e.g., Moondream) on Med-VQA/RAD for medical VQA, detection, captioning. Optimize for on-device.
  - Front-End Engineer: React UIs for patients/providers with voice and diagnostic visualization.
  - Mobile Engineer: Develop native iOS (Swift/SwiftUI) apps for seamless voice (ElevenLabs), camera capture, intents and conversational workflows.
  - Open Application for Unconventional Talent (tell us / show us? let's chat)
Lean team (5), high ownership, remote-first. Open to contracting roles.

Apply → Email tmartin@copilothealth.com with “HN [Role]”, resume / github / research. We reply fast.


> This is investigative journalism with no bosses, no advertisers, no filters.. just the raw truth about power and those it crushes.

Hard to take this statement serious when all the headlines in this substack echo leftwing bias. At least be intellectually honest about your intent.


Do you seriously think your comment is intellectually honest?


This letter sounds like it's written by an activist, not an investor.


An activist investor, hopefully for Tesla's sake


They're being illogical. Elon doesn't get paid unless the company executes, which he's already succeeded in his prior pay package. Pure irrational thinking. And much of the comments on this post reflect similar idiocy.


I was thinking the same thing, a lot of people have an interest in Tesla failing, no surprise here


Revenue is flat last three years, profits halved, margins crushed and there will be no profits without the EV credit...Valuation still priced like 2021...

Meme stock...


You forgot "CEO is pushing for massive raise".


> I became my father’s assistant, carrying props, rehearsing patter, acting as the straight man. But I was also his skeptic. If a trick fooled me, I made it my job to discover how. When he succeeded, I applauded; when I found the secret, I felt the satisfaction of uncovering a law of nature.

I find this beautiful

> My father taught me to vanish before I learned to appear. Science taught me to appear without vanishing — to stand by evidence, to let truth emerge even when it contradicted the spectacle.

Poetic


I love the simplicity of this design


Welcome to follow anytime.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: