The issue with that promise is that anyone can convert pdfs, the question is whether the conversions are correct or whether you have
Income Expenses
200 100
On one document, and
Income Expenses
20 0100
On others.
There's no shortage of products that tried to solve this problem from scratch (or by piggybacking on other projects) and called it a day without worrying about the huge problem that is quality and parseability.
The most robust players just give you the coordinates of a glyph and you are on your own: Textract, PDFBox.
Vow is a lab grown meat company taking this path. Last year they made a mammoth meatball that used mammoth dna with some elephant dna. I think this is a smart approach, allows them to charge higher prices for a unique experience. [1]
Other types of meat that make sense are high end sushi fish etc. where again the conventional price they compete against is high enough that it makes it economically possible to grow it in a lab.
Of course winter does not happen at the same time throughout the world. Would be interesting to see if it does the same if asked in a language spoken in the southern hemisphere.
That's a great question. However, I'd suggest a different approach as:
1. Southern hemisphere is mostly water
2. Majority of the text online is in English
So maybe a better approach would be to geolocate the prompt in Australia, for example: it's December in Melbourne, Australia...
December in Melbourne (and most of Australia) is a wind-down month. Over the week between Christmas and New Years many jobs essentially stop as we have three public holidays (which will always fall on week-days) across six days and people tend to take the opportunity to travel and enjoy the nicer part of summer. People just don't put in the work in December.
You can use log2ram [0] to reduce the number of writes to the SD card. That's what I'm doing at the moment and haven't had any problems with my SD card, though it has been in use for less than 6 months so far. I agree though that a SD card is not the best storage option for the OS and would like to move mine over to an nvme drive.
Safe is interesting but it’s still pretty much still a ledger. I am not sure this will work at current-internet-scale, let alone in a couple of decades worth of data. In addition to that they still pretty much keep building on top of the web, which is really not accessible to machines.
Also the problem with GMOs is it transfers ownership of the plant from the public to the large companies that have modified it (as they will inventibly want to patent it). Is it really fair that they take thousands of years of evolution, change a gene or two, and become the sole owner of the strain? And often limiting the ability of the plant to reproduce, so you have to buy more seed from them. If it continues that way the amount of plants available under public ownership will continue to decrease.
Plant patents date back to 1930, predating even the discovery that DNA is the genetic material.
In many crops the seeds planted are hybrids, which don't breed true either. That non-issue has been there since before you, and possibly your parents, were born.
What keeps the (gasp! shudder!) corporate ownership of these things from sinking us into a dystopian hellscape? Competition. You don't like one kind of hybrid? Buy a different kind. The seed makers compete fiercely for farmers' business. The same is true of GMOs.
Plant patents are considerably narrower in scope than utility patents, at least in the US. In particular you can't get plant patents that cover sexual reproduction of plants or asexual reproduction via edible tubers (like, say, potatoes do), so you can't use them to stop farmers replanting their seeds. That's why agritech companies wanted to be able to file utility patents on plants, which they eventually managed thanks to GMOs.
Also, the whole "plant patents date back to 1930" line is literally out of Monsanto's talking points.
Talking points tend to include facts. I'm sorry if the facts are not helping your narrative.
Yes, the 1930s plant patents are a bit different. But they are still corporations owning IP in plants. It didn't end the world then, it's nothing but a source of faux outrage now.
Plant patents (and patents on new GMOs) are an excellent idea, and it's remarkable you could think othewise. They mean we actually get the improvements, since they cannot be immediately taken by those who didn't invest in their creation. They do NOT mean you have to stop using something that existed before the patent.
Monsanto had a contract with farmers to not replant the seeds. The farmers were free to not use the seeds and not sign the contract. Enforcement of contractual obligations is a vital glue that holds society together; why are you against it in this case?
In all this, I detect entitled selfishness on your part: you want (or you think the farmers should have) the benefits of these improvements, without having to respect the mechanisms that allow the improvements to be made.
OK, I think we are agreed that all seed providers would use the similar kind of contract. (I'll assume you concede this, since you didn't address my question about it.) In which case any claims of "competition" and "freedom" are somewhat hollow.
For whatever it's worth, the research we're discussing here is being done not at a for-profit seed maker but at a non-profit university. Universities are good at getting government grants, for example, for innovations that would benefit everyone, like improved seeds.
There is no reason for a supplier of hybrid seeds to have this sort of contract, as the hybrid would not breed true.
As for patents: while patents in general have dubious applications, plant patents in particular are not dubious at all. Patents are absolutely essential here, because the cost of manufacturing is extremely low (just grow the plant!) If plants that breed true could not be patented then developing them would be worthless. The cost of development could not be recouped after the first sale.
Ah, I had the exact same idea for a website after seeing the post on spleeter here on hn a bit ago. Just tried an mp3 and it works great! Even though you have beat me to it I think I'll still have a go at making something similar to get some frontend experience :)
There are projects like Yggdrasil [0] and cjdns [1] which are encrypted, distributed networks. People will generally use these networks on top of the existing internet infrastructure though, because for now that is the most prevalent way computers are connected to each other. But there is nothing stopping people setting up their own mesh networks which could completly or partially avoid the existing internet infrastructure which governments have control over.
Also you may be interested in the SAFE network [2], it's not live yet but been in the works for some time and is getting close, more detail at [3].
Very interesting and exciting work. Michael Levin (one of the scientists in this study) has some amazing things going on in his lab. There is a few talks of him on YouTube, I have seen the following one from May this year
All sorts of work about setting the electrical potential between cells/parts of the body to certain states to induce the desired growth. E.g. growing a eye on the gut of a tadpoles.
May also be interested in Allen AI's OCR tool olmOCR they just released too [1][2]. They say "convert a million PDF pages for only $190 USD".
[1] https://github.com/allenai/olmocr [2] https://arxiv.org/abs/2502.18443