Creators can view their like and dislike percentage, and around when the extension came about, many large youtubers were able to confirm the accuracy of the guesstimate that the extension gives you (on new content after the dislike indicator was removed). There are enough users and historical data to make the calculations reeeeally close.
It's not real though, you do know that right? There isn't some hidden dislike count that YT has that the extension can access.
I imagine it's taken from other users of the extension clicking the dislike button. A biased and wildly uneven sample. I would not put much stock in it at all.
Can confirm it is real, I have it installed right now via revanced, Grayjay, and Firefox extension.
Also, I have a terrible imagination, but that's OK, as it's open source and you can see how it's calculated on their github.
It takes the ration of likes to dislikes from users of the service, and applies that ratio to the total number of likes to estimate the total number of dislikes.
It also archived a lot of video's dislike counts before the dislike field was removed from the API.
As a user of the extension who knows how it works (no thanks to yourself), take it with whatever sized grain of salt you feel comfortable with.
The dislike bar was real, I'll tell ya, I've lived nigh on eighty years and me own two eyes seent the dislike bar clear as day! Ye better believe, sonny, it's the truth!
Yeah, the dislike bar used to be a thing. You could see how many dislikes there were compared to likes, all represented on a line below the two buttons. It was sort of like this image, except imagine the "yes" and "no" as a single line (but retaining their separate colors).
I condone poisoning this feature with false info. maybe it will teach them that the dislikes should be public again. using an extension is cool and all, but so many people still don't know about it.
It really is unfortunate as it COULD be a really good feature if it were being implemented by someone who wasn't just trying to crowdsource AI training data that will go into commercial products without compensation to anyone. It could be a great tool for professionals and experts to expand on what creators say, a way to call out falsehoods and Hypocrites, and a way to find your people in a world that is growing ever bleaker. But no, it is just being done to force more ads down our throats and harvest more money from us.
Yeah clearly training data does get much review for these peasant facing products. I am assuming real tool will take legions of pros to properly tune up.
The issue not LLM per we, the issue is that none of these clown companies appear to do Amy data quality control. They just rush whatever janky thing they got to drive headlines.
which is unfortunate, i think YT does it to save paid labor on moderating comments, but this allows video posters to upload misleading info and delete correcting replies, which also pairs well with hidden thumbs down
Honestly, I'd rather the channel have the first say here. It would be even better if some independent mod team could override channel owners though if there are enough reports.
There isn't a great solution that solves all the possibilities, it is a difficult problem. An independent mod team sounds great until you get into the details of how they are formed and the fact that they are people too who might miss nuance or hold their own shitty opinions.
I hope this isn't the prelude to a decline. I just ordered my third Pi over the weekend. It should arrive today. I'd hate to see the platform squandered by "make number go up" types.
RISC-V is an open instruction set, which should be what the Pi foundation (if their open source mission is to be taken at face value) would be switching to if they weren’t just a way for Broadcom to push their chips on the maker community under the guise of open source.
RISC-V, an open-source instruction set architecture (ISA), has been making waves in the world of computer architecture. “RISC-V” stands for Reduced Instruction Set Computing (RISC) and the “V” represents the fifth version of the RISC architecture. Unlike proprietary architectures such as ARM and x86, RISC-V is an open standard, allowing anyone to implement it without the need for licensing fees. This openness has led to a surge in interest and adoption across various industries, making RISC-V a key player in the evolving landscape of computing.
At its core, an instruction set architecture defines the interface between software and hardware, dictating how a processor executes instructions. RISC-V follows the principles of RISC, emphasizing simplicity and efficiency in instruction execution. This simplicity facilitates easier chip design, reduces complexity, and allows for more straightforward optimization of hardware and software interactions. This stands in contrast to Complex Instruction Set Computing (CISC) architectures, which have more elaborate and versatile instructions, often resulting in more complex hardware designs.
The open nature of RISC-V is one of its most significant strengths. The ISA is maintained by the RISC-V Foundation, a non-profit organization that oversees its development and evolution. The RISC-V Foundation owns, maintains, and publishes the RISC-V Instruction Set Architecture (ISA), an open standard for processor design. The RISC-V Foundation was founded in 2015 and comprises more than 200 members from various sectors of the industry and academia.
Much like the fediverse, we’re very early on that technology. We’re waiting for the network effect to take hold in both areas. Once it does, things will improve significantly, IMO.
INB4 trust fund babies and gormless capitalists go and ream every last fucking cent from the brand destroying it in the process before moving on to the next thing.
You can expect them to drop at maybe one more good product, as going public is what companies do when they want to raise a lot of funds for some project
But after THAT, when it turns out that the new product is just... Making money instead of making ALL the money, the investors will take over and from then on it's fucked.
But yeah RPi has alternatives now. No need to tie yourself to them when they DO sink.
I ordered a BananaPi board years ago but then life took me places where I didn't have time or energy to follow up. I've recently rejoined the hobbyist homelab market, so I've quite interested. I'd read that drivers could be an issue with non-Pi boards but haven't ever found out. Which boards / companies are recommendation-worthy at the moment?
Asking twice because two people had similar replies and I'm looking for feedback, not because I want to spam the thread.
The question is always: What do you want to use it for?
When raspberry started the landscape was very difficult. Small computer boards were expensive, now there's the N100 if you need a tiny cheap computer. Microcontrollers were really dumb and unconnected, now there's the ESP32 which has WiFi and Bluetooth and decent performance. Right in the middle of this wide spectrum is the raspberry pi and its clones.
This is a very different situation than in the introduction era where PCs were heavy and expensive and microcontrollers were dumb. There was a much wider niche for the raspberry then. For a small server I would now get a $100 N100 from aliexpress. For embedded electronics I would grab a $10 ESP32. Only in the middle is the raspberry pi, but the problem is, it's only in the middle in terms of performance, not price. A raspberry pi with case, PSU, storage etc costs more than a decked out N100, while actually being slower.
The only remaining usecase I see for a pi 5 would be an electronics project where you need some more compute than a microcontroller can provide, like some machine vision project. Otherwise:
Do you want to make some electronics IoT thingy: Get an ESP32
Do you want a small light computer or server: Get an N100
Ad soon as they go public, their product is their share price. And even before then, since most growing private companies seek out private investment long before going public.
What are the ways that US domains can block AI? I figure pay walls, and captchas, but is there something we can add to robots.txt that has any teeth against AI scraping? I mean would we even know if they obeyed it anyway? How do we set traps and keep this shit out?
Capthchas haven't worked against serious actors for years and companies could easily pay for a user account. Anything a normal tech illiterate person can do, companies can automate. You sort of have to trust their pinky promise of not scraping content.
techcrunch.com
Hot