Warm hearths for heavy hearts. ml backend infra, six stringer, (over)thinker
milkyway
Joined June 2024
- Tweets156
- Following133
- Followers52
- Likes1K
ramján retweeted
建议大家千万不要一边做一边怀疑
1.只要信念极度坚定，梦想大概率都会成真。这个世界上极少有不能成就的事情。大部分的事情都是可以成就的，大部分的事情都是可以实现的，都是可以触摸的，都是可以收获的。而绝不是相反。而且是越是光明的东西，越是正大的东西，越是大道上的东西，越符合这个真理，越符合这个逻辑。因为这个世界的真相、这个宇宙的真相就是，人与他周围的任何东西，与他周围的任何人事物境，其实本质上都是一种框架的博弈关系，都是一种配得感的博弈关系。
2.我之前看一个通天术，有过。上面讲的是如果想干一件事吧。第一，先相信自己。第二，极度的相信自己。第三。直接去干。第四。排除异己，把不相信自己的人全给排除掉。只留相信自己的。最后基本都能成。因为他的核心理论就是位置决定脑子。
3.不过分纠结于状态，你任何状态都可以做事。真正的创造者不随情绪起舞。他们快乐时创造，难过时也创造。只管开始做。
4.如果一个人有强烈意愿做一件事，是一定能做成的。当然社会、环境、他人会有诸多桎梏，但是你自己要意识到，这个世界不会天生给你让路的。你需要像一艘破冰船一样，逢山开路遇水架桥，虽千万人吾往矣。一个人内心中真正强烈的蓬勃的意愿，无论是志向、兴趣亦或是欲望，会带你走向真正想去的地方。
5.不管什么事，决定了就立刻去做，这本身就能使人生气勃勃,保持一种主动和快乐的心情。
6.在大多数领域，与决心相比，天赋被高估了。时间长了，决心就是你的天赋。
7.不要将自己的主要精力放在抚平情绪上
8.人确实活一个心态、活一个能量，而不是活一些条条框框。心态好、能量高，才能扎进自己的生活里，把握自己的轴心，做成事、爱到人，享受到这世间最瑰丽的景象。
spinning up by openai is an excellent resource for RL!
spinningup.openai.com/en/lat…
spent the whole day preparing to talk about my projects for an interview, only for none of them to come up.
now i'm sitting, wondering how much better it could've gone, while also having learned the 'secret sauce' of how businesses work from the guy who interviewed me 💀
7
453
suppose you have a good old engine in your car that you've used for years. it's battered, but still kicking. you trust it, and you're used to it.
every week, you drive it from your home to the market.
now someone comes up to you and tells you they have a new kind of engine that will let you drive to that market in a much better way.
but you've always made that drive with your old engine. you're used to it, and you don't have a problem with it. so why would you replace it right?
now, suppose you're promised that with the new engine, you won't just be able to drive to that market in a better way; you'll be able to drive to a BETTER MARKET.
would you choose to replace it?
5
219
what if math is becoming more like an observational science, like astronomy, and ai is our galileo's telescope. it lets us look far, far away. we'd just have to make sense of what we see instead of exploring on our feet.
1
8
308
i kid you not, i just overheard someone literally bragging about 'orignal' first copy shoes lmao.
1
6
198
ramján retweeted
We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:
Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:
Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:
Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:
Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!
In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
ramján retweeted
One of my friend makes $1.2 million a year as an Anthropic engineer.
I asked him how he learned prompting so well. He sent me a video that was never supposed to get out. Their core team's prompting playbook.
You won’t find anything better about prompting than this 30 minutes video.
I watched it last night. Halfway through, I realized I've been using Claude completely wrong for two years.
Watch it, then read the article below.
to anyone living alone:
i recently started living by myself and I don't like it. now, there could be two things: either you eventually fall into the ebb and flow of this kind of life by habit, or you've somehow performed some voodoo mojo that I haven't figured out yet.
1
11
1,248
is it just me or do people with a lot of years of experience seem more clueless about where the industry is headed than we are?
2
8
250
