The advice about deleting youtube history and setting and auto-delete cadence (though every 3 months looked like the most frequent possibility for me) is good and I wasn't aware of it. I don't have social media, but I do have a personal gmail email address, which links to youtube making it hard to avoid spending time on my phone watching youtube videos.
It's crazy that insurance companies are (rightfully) viewed so negatively that the killing of a CEO is responded to positively in a lot of circles. No doubt the jokes on twitter have been great, and it is a good release like I've heard people say about breaking a window with a brick during a protest. It (edit: represents) a lot more than just property damage. It's unlikely but maybe some good will come of this from insurance companies.
True, but you could also frame this as a way for Anthropic to try and break that trend. IMO they've got to try and compete with OpenAI, can't just concede that OpenAI has won yet.
Is there any reason why this would work better or is needed compared to taking audio and 1. doing ASR with whisper for instance 2. applying an NER model to the transcribed text?
There are open source NER models that can identify any specified entity type (https://universal-ner.github.io/, https://github.com/urchade/GLiNER). I don't see why this WhisperNER approach would be any better than doing ASR with whisper and then applying one of these NER models.
This works better because it gives a secondary set of conditions for which the decoder (generating text) is conditioning its generation. Assume instead of their demo you are doing Speech2Text for Oncologists. Out of the Box Whisper is terrible because the words are new and rare, especially in YouTube videos. If you just run ASR through it and run NER, it will generate regular words over cancer names. Instead, if you condition generation on topical entities the generation space is constrained and performance will improve. Especially when you can tell the model what all the drug names are because you have a list (https://www.cancerresearchuk.org/about-cancer/treatment/drug...)
It would be cool to see one of those word diagrams where the size of the word is how often it appears (a word cloud), and to have one of the word clouds for the word in 1970 and one for the words in 2018 with maybe some years in between. That would make it a lot easier to digest the information than a grid of frequency line plots. It's information overload when I open the page and it takes a lot of energy to read all the different words and compare the plots. The word clouds would get the point across easier and clearer imo.
An assumption of the article is that in addition to getting readers, bloggers want to make money from their blogs via advertising, brand endorsements, etc. That's fair and true for the author in this case, but not necessarily true of all bloggers, especially the tech type that are on hacker news.
I used to publish some of my edge use case and home lab networking experiments. I had about 150 pages on L3 switches, Wireguard, IoT connectivity.
No ads, no SEO, just sharing information. By the beginning of the year, my only requests were bots. The only referrals were from someone on Reddit linking to a page.
I took it down. Google extracted the value of my work, they are doing the same to everyone else.
Isn’t that obvious? They are a misanthrope. People forget corporations act on behalf of their consumers (and democratic governments on behalf of their people). It’s not something to moan and whine about, but seems like HN is a haven for that attitude.
Fascinating, I missed a lot of that in the comments before you, but when you say They are thinking the Bad Thing for a bizarrely specific ideological reason...the straightforwardness of your reasoning gives me ability to shut my brain off. And that makes me feel very comfortable, and finally, if it feels good, it's true
He even mentions the "Blogging Apocalypse" and follows with a downwardly sloped graph titled "bye bye traffic" with no units or further explanation.
I shortened my sentences. I used keywords that Google could identify easily. I wrote in a way that allowed Google to understand our content,[...] If I wanted people to find our article on Prague in a Google search I had to call it something Google understood. And then I had to repeat what the article was about in the first 100 words. And then do it again and again in the content. It led to some less than stellar paragraphs occasionally, [...]
If the "blogging apocalypse" can rid us of SEO spam, then perhaps it's not a bad thing?
I'm surprised the article and non of the comments I've seen in this thread mention effective altruism, which in the framing of the article is an attempt to scale caring in some sense.
The article helped me realize why so many people in the startup/tech/software engineering scene are drawn to effective altruism, it's a way to scale helping humanity in the best way possible. The effective altruism argument is that it's more productive to by mosquito nets in Africa than to volunteer at a food shelter because buying mosquito nets will save more lives.
But this article helps explain why I don't personally feel drawn to or buy into effective altruism. Because caring doesn't scale, you can only really care about a handful of people close to you. And to me that altruism feels better and seems like it goes further than donating money to help more people that I can't truly care for.
Mosquito nets are oft being used as fishing nets. Since 1) the loopholes of malaria nets are much smaller than manually knotted ones (thus also killing baby fish) and 2) the impregnation chemicals kill riverbottom insects (thereby disrupting the foodchain), this has lead to a massive killing of fish. Instead of fixing malaria, this has instead led to aggrevated hunger crises in some places.
I think most people's problem with "effective altruism" is rather that some of its more prominent adherents end up defining "helping humanity" in very weird and rather counter-intuitive ways. Like that whole argument that we basically should care more about future humans than present ones because there's a lot more future ones.
I agree that's most people's problem, but it's now mine. I've tried to think about effective altruism from first principals. For me, I can rationally understand effective altruism and why it's attractive. I even believe people that follow it (and actually do good) may in some sense be "better" people than me. But it none the less rubs me the wrong way and I can't follow it, so I've tried to understand why.
