

I see you don’t run electron app in flatpaks :)


I see you don’t run electron app in flatpaks :)


Unlike fern and tldrnews, I don’t think they declared no AI, but I feel most of there animation seems to involve a lot of human labors, at least on top of AI.
They have also never declared the use of AI either, so I guess I don’t know for sure.


Veritasium
Fren
Johnny Harris
Compterphile
3blue1brown
tldrNews (several channel each for different region)
RealLifeLore
Money and Macro (actual economist, not finance bro)


I teach computer science at a uni. I would much much perfer student to ask me the question instead of LLM. I will give them a much better answer because the LLM probably learned from my published stuff, which is already super surface level, and then they will distill again.
I understand this is not feasible for larger classes, especially introduction classes. But please, if you ever move on to upper level classes and professor welcomes questions, talk to the professor, not the LLM.


I got a lot of flipboard content on mastodon, but I don’t know if that “grows” flipboard.


I find the social media (piefed and mastodon) I use are busy enough to keep me entertained, not too busy to make me addictated.


BTW, it loads fine now :)


I am able to access it pretty consistently on interstellar and the web, but it is so slow it is not really a decent experience.
For me, connecting from the U.S., takes around 5 – 15mins, but I can barely vote on anything.
I remember there is a post stating that LemmyNSFW is eating up the VPS resource, and the provider throttled the VPS, which also include FediNSFW. My guess is that after LemmyNSFW is down, people are eagar to backup that instance, which created an absurd amount of traffic.
What about when you hang out outside on a weekend, then that is easily two meals outside.
Do you not hang out outside that long or do you go home for lunch and go back out?


For additional reference, boiling a liter of water takes just shy of 100Wh (not counting heat absorbed by kettles or heat distributed into the air): https://www.quora.com/How-many-watts-do-I-need-to-heat-1-liter-of-water


As far as I know it is still much more expensive compare to alternatives like grammar checker and web search. Especially give that model already searches the web on its own in many queries.
Just because inference is more efficient than an training, which consumes energy on the scale of nation states, doesn’t mean inference itself is econonical.


I used ddg and startpage, they are both great. Although I slightly perfer ddg, but stays with startpage because it is european.
I also turn on ads to support them. However, now I have more money but less time to scroll pass all these ads, I switched to kagi, which is ad-free and fast. The price is not cheap and they are not based in Europe (they are registered in the U.S. with employees all over the world, which is technically better than U.S. centric ddg), however there is no alternatives that I am aware of.


I never talked to my parent about this, but I imagine way better. My dad is one of the earlier master degree holder in China. He got his PhD much later; I believe around his 40s.
At my age, my dad is an entry level doctor and my mom an entry level nurse. They live in a room of 10 square meter assigned by the hospital they work in.
I am at the end of my 20s and is an assistant professor; a very busy job, but nevertheless stable, which is an excellent perk in this economy. The pay is not bad (not as much as industry though) and the work is really fulfilling. The only unfortunate part is that I almost always need to work overtime. It is not uncommon for me to work from waking up to going to bed, if not working into midnight.
LLM is very good at programming when there are huge number of guardrails against them. For example, exploit testing is a great usecase because getting a shell is getting a shell.
They kind of acts as a smarter version of infinite monkey that can try and iterate much more efficiently than human does.
On the other hand, in tasks that requires creativity, architecture, and projects without guard rail, they tend to do a terrible job, and often yielding solution that is more convoluted than it needs to be or just plain old incorrect.
I find it is yet another replacement for “pure labor”, where the most unintelligent part of programming, i.e. writing the code, is automated away. While I will still write code from scratch when I am trying to learn, I likely will be able automate some code writing, if I know exactly how to implement it in my head, and I also have access to plenty of testing to gaurentee correctness.