ZeroCool2u

This is handy, but I wonder how much it will cannibalize their services. I have a simple app deployed on cloudflare for a very niche single purpose use, but I wouldn't have bothered if I had this. Serving it from my own machine would have been fine.

Anything specific to watch out for?

If you enjoy DCS give VTOLVR a shot as well. Fantastic fun.

https://store.steampowered.com/app/667970/VTOL_VR/

You can pick up a Valve Index if you want to do this, which is a fantastic device, but I think the reality is that the market for the wired to a desktop head set just isn't there to support the product.

I'd be interested in using this with something like the new HuggingFace robots. Do you have demos showing how to go from HFlow dataset to training to deploy?

The requests packages fundamental API is good enough and is now maintained by the PSF. I don't understand why OpenAI and Anthropic don't just throw some $$$ to the PSF for focused improvements and expansion of it.

I have 5 Gbps symmetric fiber in NYC with a Google Fi data only sim as a built-in failover to 5G on my router. It did require some specialized hardware, but it's truly a wonder coming from a kid that experienced dial up.

It really changes the calculus on some things with large volumes of data on my desktop. Admittedly, my desktop maxes out at 2.5 Gbps, but everything else in the house keeps happily chugging along.

I'm shocked Teams + GitHub haven't already done something like this. But I suppose that's asking a lot from Teams.

I've been using Claude and other agents in Slack for some code tasks. It's mostly fine for simpler projects and products. Things like docs or specific libraries. But I'm not really convinced this is the ideal interface yet. Reviews are my bottleneck and GitHub has a great UI for this. Plus, token consumption with Slack is out of control.

Maybe in time this will be the way we work, but for now I still find myself using GitHub, a terminal, and slack together.

I don't disagree about distribution, but absolutely no one I know outside the tech bubble knows what Lovable is or has even heard of it.

It's our guy Daniel: https://www.linkedin.com/in/danielhanchen

Unsloth already has a guide for their quants: https://unsloth.ai/docs/models/qwen3.8

I transitioned into a Product role this year, coming from a CS/SWE/DS background and it is just shocking how poorly the AI models perform at written content.

Domino Data Lab | Remote (US) | Multiple Roles

Domino is hiring for a number of roles!

Great pay, hours, culture, product, and mission. I went from being a long time customer to being an employee, because of my conviction.

https://domino.ai/careers

I have this exact same problem. It's really the only annoying thing about my apartment. Would absolutely pay for an off the shelf solution to this problem.

Really cool! Just ordered one. Do you guys have an active lab in SB?

I used it for double checking some stuff while helping my friend build his PC! I had it in tight spaces and it helped me verify some details around which nvme slot to use first.

Gemini live has been able to do this for over a year now. I can just activate it on my phone and it really works surprisingly well, especially the interruption. I've tested it with my 95 year old Dutch grandmother and it switched seamlessly between English and Dutch with her and handled her poor hearing very well, including her asking for repetition.

I'm a little surprised by how much OAI is playing catch up here.

The tutorial is helpful: https://model-checking.github.io/kani/kani-tutorial.html

Reminds me a bit of hypothesis auto in its simplest applications: https://github.com/timothycrosley/hypothesis-auto/

Having used it quite a bit when it was out, it's not. It's certainly better, but in some ways it's worse. It's trained to be more "agentic" and even in cases where I wanted to talk things through first and I would explicitly tell it not to do something, it would take action on my behalf without checking first.

It's also still just prone to the kind of "stupid" mistakes we see from all LLM's. Like it can write great code, but it doesn't really have common sense without enormous guidance.

This one? https://www.cs.huji.ac.il/~nati/PAPERS/lmn.pdf

More information here: https://youtu.be/66QzlDewigE?is=WC9vt6LBtdWFPqfp

One can only hope The Culture is the trajectory the future bends to.

Zed is excellent. I know it's weird, but the last thing holding me back is being able to have a browser based Zed session the same as VSCode.

A crucial factor tech industry folks tend to ignore is how much executives value predictable costs. Cloud migrations got away with this, but still had to argue fiercely, because 'the cloud' and its serverless tech had the potential to significantly decrease overall spend for unpredictable, bursty workloads.

The usual counter-argument is the operational burden, but human capital is also a relatively fixed cost. A dedicated team of 3-5 FTEs could probably handle inference ops for a F500 company.

Meanwhile, the capability delta is shrinking fast. We have more evidence that local open-source is viable with the release of DeepSeek v4, and the industry is only trending further in this direction. Especially as we rely more on test-time compute and task-specific harnesses rather than model size.

So, if you're an executive looking at a marginal but fixed operations cost, added flexibility, and a rapidly closing gap in capability, why wouldn't you just run open-source models on your own infrastructure to get those highly predictable costs? Plus, you decrease the risk of one of the frontier

Interestingly, Anthropic uses Mintlify for their docs. Not Stainless. Obviously, the focus is on SDK generation, but still strange.

It is frustrating, because I really enjoyed my Valve Index and want a replacement and Meta has some of the best VR tech in the world, but I've waited 6 years for Valve to release their new headset to buy a replacement, simply because Meta can't be trusted.

The quality of apps in the Google Play Store has dropped massively. There are still some gems, but for better or worse, the ecosystem is simply not as strong as Apples and it's certainly not comparable to just having a device where you can install anything you'd like in a full desktop grade OS.

There was a time where Google could've been competitive in this space, specifically against Apples MacBook product line, but that has long since passed. The 3rd party manufacturer path means Google isn't committed to this and won't have competitive hardware. It'll just be another Chromebook and limited to the Google Play Store too, which just isn't good at this point.

Bedrock is both more expensive, less feature complete, and less reliable in terms of raw volume of 500 errors.

Interesting side effect of this is that Google Cloud may now be the only hype scaler that can resell all 3 of the labs models? Maybe I'm misinterpreting this, but that would be a notable development, and I don't see why Google would allow Gemini to be resold through any of the other cloud providers.

Might really increase the utility of those GCP credits.