Hacker News new | past | comments | ask | show | jobs | submit | bilater's comments login

I've noticed a lot of people from the middle of America who grew up lower middle class are stingy (even if they are doing well now). Small things like lowering the heating in their house to uncomfortable levels to save 10 bucks a month, not buying a round of drinks. etc

It's interesting because ppl in even worse socioeconomic situations in South Asia are the opposite and go above and beyond to be generous. I guess its because being hospitable is a big part of the culture there.


This sounds more like the specific cohort that experienced the Great Depression.

The whole prescription business is such a grift. Doctors deliberately give you max 2 years of a prescription that you know you will need for longer (such as allergy meds, eye prescriptions) and then you have to waste your time and spend money just to renew it.


Is this part of the reason those online pharmacy platforms (which include a doctor who will do a very quick exam/diagnosis) took off in recent years? I mostly heard about them with regard to ADHD meds but wonder if they are also useful because they don't require you to have checkups all the time for maintenance medication.


Most of those platforms seem to require the patient to pay for a virtual visit for every refill, or require some kind of a monthly fee.


That is accurate. But those virtual visits are 50% cheaper than full telemedicine visits and 70% cheaper than in-person renews.

It should be noted their virtual visits are mostly performative. You're filling in a survey beforehand and the doctor is just there to read the survey and push a big green button.


How long do the refills last? I would think that even with the cheaper cost of medicine, and the discounted telemedicine visit, it would still be more expensive than seeing a doctor annually and getting refills for all prescriptions for the rest of the year without needing additional appointments. But perhaps I'm underestimating how cheap these platforms are (both the meds and the docs)?


They aren't that cheap. But when your doctors appointment is $120+/each, they don't need to be that cheap, just cheaper. They're like 20% cheaper all-in compared to the traditional way, and you don't need to leave the house.

Plus many of these places specialize in stigmatized products like anti-bolding, ED meds, anti-depressants, etc.


When I need a prescription refilled I request a refill from the pharmacy. If the prescription has expired the pharmacy calls or faxes my doctor's office and they send a new prescription to the pharmacy. There is no charge for this.

I've only ever had doctors that worked that way so didn't even know that there were some that did not.

I now wonder which is more common, and if there is some way to tell beforehand when choosing a new doctor which kind they are?


Agreed. I have been on the same generic blood pressure medication for ~15 years now. I was recently on a vacation in Mexico and surprised to see that you can just buy it off the shelf there. That being said my healthcare plan includes a free yearly checkup with my doctor so it's kind of nice to have a yearly baseline of bloodwork/labs and be able to see it my patient portal.


Many (most) antihypertensives need periodic monitoring for electrolytes or kidney function. In Mexico, they are not going to get sued if you go into renal failure as compared to the US. Partly done for you and partly because of the sorry tort system in the US


It's super obvious from his writing his first language is English. Most likely Hal Finney.


I used to think so too, but the case for Len Sassaman is similarly strong.


What is the strongest evidence AGAINST Sassaman?


His wife doesn't think it was him (https://twitter.com/maradydd/status/1364325186372304904?t=yY...) and the original Bitcoin code/executable was Windows based whereas Sassaman was known to Mac/Linux


Those are weak pieces of evidence.

Meredith appears to be telling the truth. She didn't say Len wasn't Satoshi, she simply said to the best of her knowledge he wasn't. That doesn't mean he wasn't working on it covertly.

One of Len's best friends (Bram Cohen) knew Len was posting pseudonymously on the cypherpunk mailing list but never knew what handle he was using. Also, when Bram was about to release BitTorrent Len tried to convince him to do it anonymously. It's not hard to believe that Len would have done it secretly; even from his wife.

Furthermore, Meredith can't be 100% trusted. When Satoshi handed the project over to the maintainers and stopped posting to the cypherpunk mailing list in late 2010, Meredith tweeted, "Bitcoin isn't ready for prime time yet, according to its creator. Interested people can help finish it, though!"[1]

Satoshi never said those words publicly or privately-- so it's a curious thing to say.

As for the computer...

It's likely Len used university computer(s) for the development as the commit times and communication line up with an academic schedule. It's likely the university had Windows computers. Plus; it's one way to isolate the environment and reduce the chance of information leakage (could have even been a Windows VM).

[1]: https://twitter.com/maradydd/status/12163582133276672?t=dk8C...


The C++ coding style is also very much the way a Windows developer would write C++ code. Not the way a Unix-y C++ developer writing on Windows would write.

I always figured based on the code and the emails that it was an older Japanese developer. I've emailed 100s of them over the years and they do all typically write similar in English, they mix a lot of UK/US-ism, and often their English is really good, like I wouldn't know they weren't a native English-speaker until I just caught on to how they wrote. (Speaking is an entirely different issue, many of them cannot speak English in person very well or make obvious grammatical mistakes they don't make when typing.)

Windows is also very pervasive among developers.


Coding is quite subjective. When I examined the early bitcoin code (the one Satoshi wrote and shared). The C++ code looked pretty sloppy and amateur-ish.

The comments were odd and not standardized (randomly using four //// sometimes, etc). The use of 4-6 random new lines between sections of code was awkward. The way the code was organized, folders named, etc.

The code itself was a mix of hungarian-isms. It felt very academic-y to me... like someone that did most of their coding in university as a teacher or phd candidate (little real-world coding).

There's a podcast (name slips my mind...) where the host asked Bram Cohen if he thought Len was Satoshi and he doesn't outright say yes... To paraphrase, he basically answered, "I can't say for sure. It seemed like he (Len) lacked the C++ knowledge.. but his programming got a lot better since I last seen it... so I don't know. It seems to be the most likely scenario would be Len doing the brain work and someone like Hal doing the coding."

But, isn't that sorta what happened? Satoshi had 169 commits and Hal basically took over and cleaned everything up. Satoshi didn't do that much coding, and the coding he did do was done over 1.5-2 years (as he stated in the cypherpunk mailing list).


I'm curious, what is the stylistic difference between Unix and Windows c++? I don't know it.


Yes, the university had Windows computers... and they also had the FT in the building where Len worked


Yet she tweeted about Bitcoin in 2010! if true seems like a hugely important evidence.


Also, from what I can see, he didn't use two spaces after a period.


The biggest thing that bugs me about the Len Sassaman theory is that the original bitcoin paper, while amazing, doesn't seem like it was written by a PhD candidate. It cites just eight related works, and the W. Dai citation was famously added only after someone else suggested it, because Satoshi reportedly didn't know about it. A typical paper by a PhD candidate will cite dozens of related works, and Len Sassman would certainly have knowledge of dozens of related works off the top of his head. There should be a whole section citing the literature and enumerating the ways in which the proposals and findings in the paper are novel contributions.


They style of dissertations is dictated by the program not the authors style.


I'm convinced there is a gold mine sitting right in front of us ready to be picked by someone who can intelligently combine web scraping knowledge with LLMs e.g. scrape data, feed it into LLMs do get insights in an automated fashion. I don't know exactly what the final manifestation looks like but its there and will be super obvious when someone does it.


I feel that the more immediate and impactful opportunity that people are doing is instead of scraping to get/understand content. LLM agents can just interactively navigate websites and perform actions. Parsing/Scraping can be brittle with changes, but an LLM agent to perform an action can just follow steps to search, click on results, and navigate like a human would


Are you aware of any projects for this? I began to build my own but quickly saw that the context window is not large enough to hold the DOM of many websites. I began to strip unnecessary things from the DOM but it became a bit of a slog. L


I tried that. Turns out that LLM-generated regex is still better (and a lot faster) than using an LLM directly.


I really wish there would be more industry standard solutions like this. There are lots of semi serious apps I build where I'd rather not worry about auth and storage but its part of the product so I'll do some hacky stuff with local storage (which only goes so far)


The skepticism in the comments makes me even more bullish for this revolution.


Exponential progress. Hard to wrap your mind around.

An analogy: If it takes 20 years to create AGI equivalent to the village idiot. It take another couple of hours to go from that to Einstein.


I don't disagree, as in that if we achieve school math in a way that is not mere overfitting over a language-based training set, going to more advanced mathematics is definitely conceivable. My problem is that people are talking about solving unsolved conjectures while we are not even in a point where we know how to tackle math at all.

Imo we are not currently in the beginning of an exponential curve re solving math with AI, and def not on the path of AGI. I understand that if one believes that we are on the path to AGI soon then we shall have these math-AI advancements quite soon, but I disagree with the premise.


Nice - The mute alone is worth it


Good idea! Mistral 1B is supported by transformers.js now so I can add it!

The Text to Video model is a pipeline Text to Speech + FFMPEG Magic to stitch together a video.


What’s Mistral 1B? I only know of Mistral 7B.


I think it might be a < 1B fine tuned model. Read about it in this release

https://twitter.com/xenovacom/status/1722661501180256311


The announcement seems somewhat disingenuous. The PR[1] found from their release notes[2] seems to contain only boilerplate and no real support for Mistral models or their weights.

[1]: https://github.com/xenova/transformers.js/pull/379 [2]: https://github.com/xenova/transformers.js/releases/tag/2.8.0


The other tools section is just extra helpful links to other stuff I've done.


None of these are LLMs. LLM means Large Language Model (for text generation, like ChatGPT).


Looks like we've hit full kleenex on terms for models, maybe.

Even if it were a tiny LLM, which there are none on that site, it would technically be a SLM, for small language model.


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

Search: