z3ugma

It doesn't come with any RAM? You have to add $140 for 8GB or "bring your own"...so the list price does not represent a working computer?

Love it! Do you have a BAA with Claude though? Otherwise, your demo is likely exposing PHI to 3rd parties and exposing you to risk related to HIPAA

not re-implement SQLite, I mean "use SQLite as your persistence layer in your program"

e.g. worry about what makes your app unique. Data storage is not what makes your app unique. Outsource thinking about that to SQLite

but it's so trivial to implement SQLite, in almost any app or language...there are sufficient ORMs to do the joins if you don't like working with SQL directly...the B-trees are built in and you don't need to reason about binary search, and your app doesn't have 300% test coverage with fuzzing like SQLite does

you should be squashing bugs related to your business logic, not core data storage. Local data storage on your one horizontally-scaling box is a solved problem using SQLite. Not to mention atomic backups?

At some point, don't you just end up making a low-quality, poorly-tested reinvention of SQLite by doing this and adding features?

Didn't they already have one of these? Like the original FitBit was a screenless fitness band...

You did all of this in one week?! Pretty cool. Even with LLM-assisted making, it shows that you have a lot of taste and architecture know-how. What would you attribute to being able to think through useful abstracts like the design tokens?

I have been using this library for a few months alongside Gemini 3.1 Fast

It's really useful to get an iteration loop going with an LLM.

The OCCP viewer extension for VS Code helps make sure you can see and manipulate the resulting model

See also the recently-posted open source project https://github.com/ForestHubAI/boardsmith

"Text prompt → KiCad schematic, BOM, and firmware. No templates — real wired circuits with computed values."

Do you know the meme about carcinization ...how in nature, everything tends toward becoming a crab?

I think we are reinventing HTML from first principles. It's semantic structuring with a meaningful render

"Well There's Your Problem" on the collapse of the St Francis Dam, mentioned in Grady's video https://www.youtube.com/watch?v=hxLgM1vnuUA

Also I love when they refer to it as the "_First_ California Water Wars" in a grim realization of the future of water scarcity in the West

I will say I came upon this same design pattern to make all my chats into semantic Markdown that is backward compatible with markdown. I did:

````assistant

<Short Summary title>

gemini/3.1-pro - 20260319T050611Z

Response from the assistant

````

with a similar block for tool calling This can be parsed semantically as part of the conversation but also is rendered as regular Markdown code block when needed

Helps me keep AI chats on the filesystem, as a valid document, but also add some more semantic meaning atop of Markdown

Nah read my own comment history. I tend to phrase things this way a lot too because it positions it as opinion/curiosity rather than arrogant / confident statement?

This is such a well-written essay. Every line revealed the answer to the immediate question I had just thought of

Building a PCB milling software tool that combines converting Gerbers to toolpaths in G-Code and then actually commanding and operating a restored PCB mill from 1998.

Bringing back the hobbyist self-made PCB workflow since it can be a headache to get designs back from PCBWay / JLC between customers, tariffs, shipping..get more quick to close the prototyping loop.

Here's a YouTube playlist of how to make PCBs at home for anyone who wants a deep dive. https://www.youtube.com/watch?v=Hm_JrACrmVs&list=PLWDQgxl-jH...

Do you want some more science YouTube recommendations to top that off?

Right click > "Paste from Markdown" instead of just straight up pasting in

"78 incidents in last 90 days" per https://mrshu.github.io/github-statuses/

that's....gobsmacking...I knew it was memeably bad but I had no idea it was going so badly

There's a nice middle ground here, Software as a Medical Device (SaMD) which also goes through 510(k) in the US and has its own regulatory pathways distinct from hardware medical devices

I have a calendar reminder on August 4, 2026 to send a message to my high school English teacher from 20 years ago, for having this date stuck in my head that whole time.

This has been one of the things their latest clinical trials are pursuing, which is to see if the side effect of insomnia ends up outweighing the benefits of preventing the apnea

It's exciting seeing open-source sleep models catch up with the industry. In case you're interested in a deep-dive, a lot of commercial software like this falls under the FDA medical device product code "MNR" and "OLZ"

https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfTPLC/tp...

https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfTPLC/tp...

EnsoData | healthcare, sleep, machine learning | Hybrid or Remote (Madison, WI) | Product Manager - Software Medical Devices SaaS | Full Time | $110k-$130k + bonus, stock options | https://apply.workable.com/ensodata/j/5FD6175FEB/

Sleep is vital to recharging mentally, restoring our bodies physically, and fighting off disease. Sleep is so core to our health that it's the activity we all spend most of our life doing. Yet sleep disorders are some of the most underdiagnosed health conditions on the planet, with an estimated 80% of obstructive sleep apnea patients undiagnosed and untreated

EnsoData uses machine learning, integration with pulse oximeter hardware, mobile apps, and clinician review UIs to automate sleep study scoring and give sleep clinicians tools to find and help patients with sleep disorders.

We're hiring a product manager to uncover what features our customers crave most, conduct discovery research in the sleep industry, define and prioritize product requirements, partner with engineers to ship features, and work with marketers to tell compelling product stories. You'll manage the product development process from whiteboard mockups tested with real users through to detailed specs and launch, breaking down ambiguous problems into releasable slices that bring AI-powered sleep diagnosis to more patients.

and yet:

When you ask an AI like ChatGPT a question, what is it actually doing?

Survey of 2,301 American adults (August 1-6, 2025)

- Looking up the exact answer in a database: 45%

- Predicting what words come next based on learned patterns: 28%

- Running a script full of prewritten chat responses: 21%

- Having a human in the background write an answer: 6%

Source: Searchlight Institute

most survey respondents don't even _understand_ what AI is doing, so I am a bit skeptical to trust their opinions on whether it will cause harm

No surprises, many of these are fawning, big visions for AI infrastructure. I would love to see success from some of these less-ambitious hits that have a real chance at being successful in the next 18 months.

The "creative tools go multimodal" strikes me as one of those

Really excited for this, it might be the model that gets me to fully commit to a Starbook

Making the Nest 2nd gen thermostat the Google recently bricked compatible with local setups like Home Assistant.

I'm involved in 3 projects that are solving this problem from different angles:

https://sett.homes/

https://github.com/codykociemba/NoLongerEvil-Thermostat

https://github.com/cuckoo-nest

I'm reasonably excited about the prospect of this authorship of replacement firmware

I think that putting MQTT on this would be an important step toward local control and connecting it to Home Assistant

After you flash the exploit and SSH into the thermostat you can see it at https://github.com/codykociemba/NoLongerEvil-Thermostat/issu...

It's a boot script called /bin/nolongerevil.sh that supplies its own trust material and redirects traffic intended for frontdoor.nest.com to a hard-coded IP 15.204.110.215. 99.9% of this image is the original copyrighted Nest image. Maybe it's enough for the bounty though? And I suppose you could change that IP to a local server. If you wanted to publish the server side Nest API discovered through WireShark . Just stand up your own http rest server.

I'm working on https://sett.homes which is in this spirit. Instead of an Atmel it's an ESP32.

It uses MQTT with Wifi as you requested :)