While I agree, there's plenty of people who refuse to watch anything that's not sharp. I think there's room for both to exist, just clearly labeled as "original" and "AI Upscaled to 4K"
I used to work for a drywall manufacturer who still owned their own mines despite efforts to divest from them by some. They always viewed it as a structural advantage to still own them and not be wholly dependent on the coal plants (which effectively have conveyor belts going from the coal plants to the wallboard plants). I imagine as time goes on it'll become even more of an advantage for them to still own those mines as their competitors are forced to buy at highly inflated prices (or even from them) as coal shuts down.
Coal is expensive, but it's still cheaper than nuke and peaker thermal. If you want power fast and your state and federal government aren't worried about a few pesky environmental regulations you might see coal come back. Part of coals attraction is that it takes a lot of people to run a coal plant, and people need jobs. Those people vote and politicians like votes.
Then you would plunk down a gas turbine like everyone else. It's so much cheaper than coal to operate and uses mostly the same high capex / long-lead machinery. I could see the jobs program angle, but these are shitty jobs. It's not like working in an air conditioned mcdonald's. Workers die in mines a lot and when they don't they live shorter, less comfortable lives with disease.
I'm fine with arguing against coal for environmental reasons, but that won't convince anyone who isn't already convinced. It's always worth pointing out that gas turbines put out a lot less pollutants than coal.
Modern coal mining isn't that bad of a gig, especially surface mining (which a lot of coal is). I would certainly rather make a decent middle class wage hauling coal and support my family than work in an 'Air conditioned mcdonalds' and barely subsist in poverty.
You can make all the technical and environmental point you want. They are valid and they are largely irrelevant, at least for the purpose of achieving your stated outcome.
People want to be able to live a life with some amount of dignity and we've been so diligently eroding their ability to do so for the last 50 years that it's becoming an existential issue.
Jobs matter. If you want social progress, environmental progress, any kind of progress people need to be able to build a life where their children are better off than they were. Full stop.
Gas turbines require infrastructure the may not exist in the area yet, and significant capital outlays.
Like coal mining jobs or the like, if you’re stuck in Appalachia with 5 kids and it’s the only thing keeping you afloat, you’ll get pretty worked up if someone tells you ‘just don’t do that, duh’.
Solar is cheap and abundant during daytime. It has zero power at night. Coal/gas/nukes are more expensive, but runs 24/7. Batteries are getting cheaper, but are still not that cheap.
In areas where there are gas transmission lines, sure. Large portions of the country don’t have that infrastructure built out yet, but do have rail lines which provide coal.
It’s also a timing/capital issue.
It will change eventually, but in the meantime people need their kWh.
I’ve never seem any analysis of feasibility for LNG (probably what you’re referring too) vs pipeline NG for things like power plants. Do we even have sufficient liquifaction facilities for that type of volume? When I hear of that kind of thing, it’s almost always liquifying for export to places like Europe.
Batteries make that less true today than in the past. Solar/wind + batteries are becoming cost-competitive with coal for energy on most days (24x7).
24x7 every day of the year is much harder, though. Solar/wind + batteries are nowhere near cost competitive for reliability, though. You'd have to build a ridiculous (read: very uneconomical) multiple of typical battery capacity to make it through the long, cloudy, low-wind periods in the winter.
On the bright side, enhanced geothermal is starting to look like it may be economically competitive in the near future. If it pans out, it could repurpose a lot of the technology and labor force from the oil and gas industry to instead produce clean power. And who knows—maybe the current nuclear push will pan out and we'll have another option for reliable base load.
When Roomba thought it was about to be acquired by Amazon, it did lay off 10% of its staff - https://www.therobotreport.com/irobot-laying-off-10-of-staff.... and after the deal was canceled, it was disclosed that they had reduced R&D and focused on margin improvements, and there was some brain drain as people left Roomba as it was in a 18 month limbo - https://www.verdict.co.uk/irobot-to-cut-over-a-third-of-its-.... And of course all this self inflicted pain only hurt them doubly as the Amazon deal fell through. If they had acted as if they weren't going to be acquired they might be fine, but they tried to maximize the shareholder revenue.
Back in the day (about 2002) I was working at an education software company which was trying to get itself acquired by Microsoft. MSFT came in and told us our software didn't conform to all these "standards" in the educational software space. Standards which, coincidentally, Microsoft themselves had written. These pseudo-standards did absolutely nothing to help our customers, and were pure bureaucracy and very very complicated to implement.
I'd recently read Charles Ferguson's book about how his company was acquired by MSFT, and recognized this part of their standard operating procedure, along with extreme and invasive due diligence where they spend a lot of time working out if you're stupid/pliable enough to jump through these hoops while buying themselves time to work out if they can clone your product. I tried to warn management (yes, really - even bought them copies of the book) but naturally no one would listen, and reading a book was too much like hard work. At some point MSFT simply ceased returning management's calls, and rolled out a similar product a while later.
The company imploded not long after, not for this reason in particular, but it was part of a general pattern of incompetence and mismanagement.
Friend of mine was in a company that was going to be acquired by $bigcompany. They strung them along and strung the along until their VC funding was exhausted, then picked up the remains for a song. Much cheaper than actually buying them up.
The best open source OCR model for handwriting in my experience is surya-v2 or nougat, really depends on the docs which is better, each got about 90% accuracy (cosine similarity) in my tests. I have not tried Deepseek-OCR, but mean to at some point.
I live in a very very good area of Brooklyn and still regularly run into needles, human shit, and open fentanyl use.
LA is similar unless you never leave your little neighborhood.
DC was similar when I lived there about 4 years ago.
SF is cleaning up, but I’ve regularly walked on streets where it’s just bodies and needles
I was shocked by the Vietnamese area of Seattle. It felt like a zombie land.
I mean, if we’re talking city core yeah this it the average experience. I say this as someone who loves cities, American cities leave a lot to be desired and a lot of that comes from simply refusing to enforce basic laws that the rest of the world (including much more left countries) don’t hesitate to do.
In what "very very good area of Brooklyn" are you regularly encountering needles?!
I've lived in the NYC metro area for nearly two decades and have yet to see a single one. Definitely saw them when I lived in Baltimore, and have seen them in Philly, but even then not "regularly" in either case.
I looked into this a bit earlier this year. I'm mixed on it. While the FOSS in me wants it all open-source and available to use given that I'm basically labeling training data for them for free, and they are funded by donations/grants, I get value out of it for free.
My desire was to combine something like iNaturalist with BirdWeather for a bird tracker of audio and visual. BirdWeather does make it free which is great, but there's no great free API of iNaturalist quality for diverse bird tracking.
That being said, I am certain that if iNaturaist made their model public, tons of competitive apps would spring up and it'd be commercialized regardless of license immediately and would take people away from iNaturalist without giving iNaturalist anything in return.
Plus I know iNaturalist has issues with that they don't want autolabeled data uploaded as matched. They only want manually labeled data, which opening the API I'm sure would flood their server with ML labeled data. Which on the one hand, could be useful, but also a ton of noise.
I'm in favor of whatever option is most in line with keeping a long term success of a free, high quality plant/animal identifying app out there, and I don't know enough to take a definitive stance on that, and unfortunately those that do, probably have a vested interest in one of the outcomes.
To suggest another "simple" example, Air Conditioning. It made half the world vastly more livable, and now anywhere in the world you could work every day of the year, reduced deaths and disease. At least currently, AC has had a greater impact on humanity than AI has.
>This week, US District Judge Cormac Carney of the US District Court of the Central District of California decided that there's reason to believe that Visa knowingly processed payments that allowed MindGeek to monetize "a substantial amount of child porn." To decide, the court wants to know much more about Visa's involvement, calling for more evidence of legal harms caused during a jurisdictional discovery process extended through December 30, 2022.
So, I did some OCR research early last year, that didn't include any VLMs, on some 1960s era English scanned documents with a mix of typed and handwritten (about 80/20), and here's what I found (in terms of cosine similarity):
Handwritten is a weighted average of Handwritten and typed, I also did Jaccard and Levenshtein distance, but the results were similar enough that just leaving them out for sake of space.
Overall, of you want the best, if you're an enterprise, just use whatever AWS/GCP/Azure you're on, if you're an individual, pick between those. While some of the Open Source solutions do quite well, surya took 188 seconds to process 88 pages on my RTX 3080, while the cloud ones were a few seconds to upload the docs and download them all. But if you do want open source, seriously consider surya, tesseract, and nougat depending on your needs. Surya is the best overall, while nougat was pretty good at handwriting. Tesseract is just blazingly fast, from 121-200 seconds depending on using the tessdata-fast or best, but that's CPU based and it's trivially parallelizeable, and on my 5950X using all the cores, took only 10 seconds to run through all 88 pages.
But really, you need to generate some of your own sample test data/examples and run them through the models to see what's best. Given frankly how little this paper tested, I really should redo my study, add VLMs, and write a small blog/paper, been meaning to for years now.
