choilive

If Spotify doesnt work they are dead in the water. If Apple Music doesnt work, thats a rounding error.

But now you have the problem of being too specialized - Ive seen many specialists that think a problem lies within their specialty- when it does not. And how do you deal with problems that are multi-disciplinary (problems that require multiple organ systems) when you have an army of specialists that are each fighting for their own fiefdoms? When all you have is a hammer everything looks like a nail. Comes to mind.

Oh nice. I had dismissed offline/local web apps for anything decently complex for a long time because of the file system limitations. But it looks like this will finally make that practical with OPFS & WASM SQlite. Thanks for sharing.

Would have liked to see 24gb of vRAM in this segment, although that was probably reserved for the cancelled higher tier cards.

> Ultimately I’d prefer to serve every read from your device’s local storage, which is my current project.

Is that for native desktop/mobile clients only or web as well?

Doesnt directly answer the question, but nowadays there are good and free TTS that will essentially turn your entire ebook to an audiobook. (Elevenlabs Reader for example)

But wasn't that the tariffs? How would changing the name give any leverage?

What leverage did he gain exactly?

Always wanted a tool like this. Will try it out next time I need to inspect traffic of a docker container.

Since LLMs/AI are all the rage these days - I also found a similar behavior with LLMs.

If I ask a difficult recall question (usually some very niche medical knowledge) it will get it wrong or just hallucinate. But if I ask it in the form of a multiple choice question, it often gets it right. If memory works better when prompted with the information this way - I wonder how similar LLMs and humans actually encode memory?

I would bet a lot of money more than 1 engineer at nvda flagged this as a potential issue. If you were going to run this close to the safety margin, I would at minimum add current sensing on each pin.

No need to break them up, just ship them to the universe's most efficient storage mechanism - a black hole!

Industrial power consumers are billed differently - so it could be more or could be less. The bills are more complicated - industrial customers will generally pay for things like power factor, capacity, demand, time-of-use rates, tiered rates, etc. On a per kWh basis it would depend from customer to customer. If you had a factory that operated overnight for whatever reason, their electricity would be very cheap.

You can turn all of the AI features off via the settings.

Simple, but hard to push through. The often politically powerful police don't want that to happen.

I don't think this is a concern. Zed and Zeta are both open source. Fork it, self host it, whatever.

This is why proper differential diagnosis skills is important - something a layman would not understand. You must consider ALL possible conditions that your symptoms match, and then work your way from the most common to the most unlikely. AI is remarkably pretty decent at doing DDX.

Biological neurons != artificial neurons. You don't need to simulate biology to simulate intelligence. In the same way you don't need to flap wings to fly, or muscle fibers to move around.

You could also make an argument that the bulk of college degrees are not economically useful. The demand for certain kinds of education and knowledge is certainly finite - but the demand ceiling for general intelligence seems much much higher, and appears to be limited only by the cost of hiring ever smarter people.

The costs are coming down fast because of the investment and would not happen independently of the investment. The same reason for Moore's law can be generalized as the learning/experience curve. You could make the argument that the demand for AI has a higher growth rate than the learning rate ex. if AI becomes 10x more cost efficient and results in demand growing 100x, you would also need to follow with an exponential growth in inference investment.

Yes, but not just electric cars. Theres a push to move the entire electrical system of all cars from 12V to 48V to reduce the amount of low gauge wiring in a vehicle

The skin effect is an important difference (and advantage in favor of HVDC) so it is in fact a concern of AC vs DC.

gotta upgrade your tinybird plan ;)

Been using OVH for bare metal for a few years now and no major hiccups other than scheduled maintenance.

deepseek claims they are "open source" but they are not. They are open weight.

IMO a truly "open" AI model should have 3 components publicly available: the weights, the code, and the dataset.

Without all 3 the model is not reproducible. Could make the argument that the code and data are sufficient though.

Rails also has an application level connection pool for the database but as I understand a connection pooler (pgbouncer et al) between PG and the application servers is still necessary when horizontally scaling.

I wish this whole internet fad would just die already!

Depends on where you're looking.

There are usually plenty of "business" oriented hotels near airports, business parks, central business districts, convention centers, etc . (And that definitely reflects on the trip office map). Touristy areas have more tourist/traveller oriented amenities.

You know whats calm and not distracting? A notebook and pen. You can buy a LOT of decent notebooks for the price of one of the reMarkables mentioned in the article. (~30 or so?), and it will last a lot longer as well. Im starting to sound like a luddite.

The innovation here is you have a system that emits monochromatic light, and you have solar cells tuned specifically for that bandgap, plus the system is also "naturally" concentrating because the light output is incredibly bright. 3000 suns vs 500-1000 suns in typical CPV, plus they also do waste heat recycling. End-to-end efficiency of 40% is definitely feasible as advertised.