dghlsakjg

It really depends on caching and your use case.

My deepseek bill for the last 30 days was $2.91 for 354 million tokens for agentic coding. My heaviest usage month was $14.97 for a billion tokens at deepseek. If you are doing turn based agentic coding, 90ish percent of tokens or more are cache hits.

I have another use case for deepseek that does not hit caching at all, and I pay $2 for 12 million tokens every month. I could get that down by choosing slower providers, but speed matters for this use case and saving $.50 doesn’t.

Both those use cases are bursty. I can’t sit and wait for a server to chew through my requests, especially with my non-cache use case. Anything less than 200t/s output and I start looking for a different host.

As far as quants, direct from deepseek is full quant. Good providers tell you the quant. Even the budget ones typically won’t go as low as iq3, though. FP4 seems to be the lower limit of providers that I see.

I would love nothing more than for local models to be up to snuff, but for my purposes, they get blown out of the water on all fronts by api access. Your machine has a payoff measured in multiple years (assuming you can build it for that price: $300 worth of ram is not much unless you go for ddr3, and a $500 gpu is pretty minimal) and would be a really poor fit for what I’m doing. It seems to be a great fit for you, and that’s pretty awesome.

Compare it to deepseek flash or glm. Those are far more capable and far cheaper. Or compare it to a hosted version of the same model. The math becomes much harder to justify when you compare apples to apples.

If all you do is generate tokens all day, sure, host your own infra. For people that aren’t maxing out hardware, using someone else’s is a pretty economical option. $20 with deepseek flash gets you a LOT of agentic use, and as the frontier moves forward with new models, so do I.

The same way you do it with human drivers is one option. I.e. mandatory liability insurance.

I just tried it out of curiosity. It doesn't exist in their web suite, when I tried to run it using their launcher it failed to install and crashed the installer.

I can't directly refute your claims, but I can say that a program that won't install can't do any of those things for me.

I don't care if my lawyer doesn't use word. I do care if they use a buggy, vibe coded that is still in Alpha. Word processing is a solved problem. Free (beer or speech) word processing is a solved problem. Word compatible free word processing is a solved problem. Don't put my case in jeopardy because you want to use untested, alpha software.

Which parts of your bullshit do you use word for?

A large corpus of documents does not have to mean privileged info. Put all of the public documents from a case in there and ask it about a judges reasoning.

Honestly that wouldn’t be a terrible product: Pull in all of a judges written judgements or stated logic about their rulings and preemptively craft your case around their biases.

I haven’t used many of these vibe coded clones, but the ones I have used have all been significantly lacking in depth compared to what they are ripping off. I’m sure wordcraft works fine for typing a high school 5 paragraph essay (although, I don’t think high school students still do that). I would be shocked if it can handle revision history, complex redline edits, legal brief formatting, etc.

I also want my lawyer to focus on lawyering, not dealing with Microsoft contracts, trying to keep confidential data out of copilot, dealing with proprietary formats, etc.

At the end of the day, every firm needs some form of IT. Above a certain scale you are hiring IT help to manage it anyway, might as well use free software where you can over MS. It’s not as if managing MS for a large org is particularly easy.

Cloudflare pages have absurdly generous free tiers and absurdly cheap if you exceed that.

They lowered the ram allowance. You might be paying for that now if you have the full 24gb provisioned

Does the publisher of open model weights have any exposure to compliance requirements if it’s someone else using the software, and they aren’t distributing or operating there?

I'll bite: futures contracts serve a valuable social/economic function outside of their existence, and are just a commoditization/securitization of something that arises naturally out of doing business (agreeing on a future price today). While you can gamble with futures contracts, the underlying purpose is actually incredibly useful.

Prediction betting has an argument that truth discovery is economically valuable, but I don't buy it.

If prediction betting was banned, only two businesses would be affected, Kalshi and Polymarket. If futures markets were banned, many businesses would be massively affected, and almost every business would be indirectly affected.

Why would you do a whole new expensive, time-consuming training run if you can just tweak some numbers to get the result you want?

I think your use case of incorporating novel data probably calls for training, but I'm sure there are use cases where you can tweak the weights to get what you want (say I wanted to tweak the model to be less sycophantic, or found a way to shrink it by removing weights associated with a language that I don't ever use)

Donald Trump Jr. is a paid advisor and investor to Kalshi and Polymarket. Why bother betting if they will just give you cash for having the right name.

Granted, the son whose wedding you skip has got to be lower down the food chain.

I think the latest round of models is about to do to CAD the same thing that has been happening to development. I don't know if CAD designers have picked up on this. What happened to dev is probably going to happen to CAD designers soon. I've been having Opus 5.5 design parametric openscad models, and it can do some rather amazing work.

None of the neighborhoods I mentioned are suburban. If you don't recognize them, it sort of tips your hand at not knwoing the city at all, as much as calling the entire city "Amazonville". Discovery park isn't known for views, there are other places in the city that are. Ballard is not some sprawling suburban neighborhood, its typically one of the most recomended places in the city to visit. You get there by walking the lakefront bike path from SLU (which is actually "Amazonville"). Capitol Hill is a 20 minute walk from downtown. Alki isn't walkable from downtown, but you can take the passenger ferry. I don't know where you got the idea of "sprawling suburbs".

You are allowed to personally like one city over the other. It just seems silly to base your opinion on living in one and a few days in another, and then argue when people point out that it seems like you haven't really done or gone to the places that make a place interesting.

If you think Seattle is Amazonville and worse in all ways than SF, it sounds like you visited South Lake Union and no other part of the city. Amazon's footprint is pretty limited outside of SLU, which is one of the least interesting neighborhoods in the city. Downtown just south of it also sucks.

Ballard, Capitol Hill, Alki, University, Sunset Hill... The list goes on. Every single one of those neighborhoods has a completely different vibe (many of them used to be different cities). Incredible restaurants if you put in the slightest amount of effort to finding them. Tons of fun and interesting activities. The city is on a hill above the ocean (JUST like SF), the views are hard to avoid. Year round running weather. It may be that you just like SF more, and that's fine.

This is also true in many, many countries to some degree or another. The US has no legal education requirements for elected judges and even many appointed ones. Many European countries allow lay people to act as magistrates or judges in low courts and working alongside legally educated judges in higher courts.

Hell, the whole concept of a jury trial is that the most critical function of a court is reserved for people with no legal education.

The nutrition information is highly regulated in both countries. The Japanese take it a step further and regulate the marketing part of the label. Pictures must match the contents in size, shape and appearance. Juices can only show photos of fruit if they are 100% fruit juice. Etc.

Yes. That’s kind of the point. The words aren’t that far apart in etymology either. By Latin definitions the most popular (liked by the general public) candidate will be the democratically chosen (rule by the people) one.

Chronic long-term use of high amounts of caffeine (>600 mg, or two Starbucks Grande coffees) are linked with: bone density issues, sleep issues, anxiety issues, cardiovascular strain, stomach acid issues. They aren't super bad, but definitely not harmless.

I think this is a case of rare being relative to the past.

In the 90s and earlier it was hard to be in a city and not be breathing cigarette smoke pretty much constantly. Offices had ashtrays in the elevator waiting areas for people who couldn't wait to get to their office, restaurants had half the seating reserved for smokers, there were ashtrays in the armrest of every single seat on an airplane.

Its kind of crazy how much norms have changed around smoking in public. It's not "rare" to see a smoker, its that it is much rarer than it was in the past.

Like so much in today’s world, this is kinda Peter thiel’s fault.

He bankrolled the Hulk Hogan lawsuit against Gawker as revenge because they reported on him being gay. The judgement forced gawker into bankruptcy where it was scooped up by private equity. The new managers gutted the place and forced the writers and editors to publish clickbait garbage instead of doing journalism.

That’s why Gizmodo looks nothing like what you remember.

You can - literally - have the same government force the negligent CPA pay via the judiciary. That’s what a lawsuit is.

Normally it doesn’t get to that point though. CPAs do not want to be known as the person that misfiled taxes and got a court to rule them negligent. They carry insurance for errors so mostly just end up queitly making it right when there is an error.

How long is a piece of string?

From the IRS' point of view, the tax-payer is responsible for paying all of it. You can then turn around and sue to recover any of the fees/fines or penalties incurred as a result of an error made by the accountant, not the actual taxes though.

Typically you are on the hook for paying the actual taxes and penalties, but if you acted in good faith on the negligent advice of a tax professional many jurisdictions allow you to recover the costs of penalties and interest accrued as a result of the negligent advice.

Or is that a reflection of Kia buyers...

Those retina Mac’s are intel, no?

You could just run any other OS if any of the free OS upgrades that Apple provides aren’t good enough.

FYI: it is no longer legal to make you wait for the drug dog in the United States.

They must conclude the stop in a reasonable amount of time, without delay, and if the drug dog hasn’t arrived in the time that it takes them to write you a ticket, you can not be detained any longer.