HNHacker News
TopNewBestAskShowJobs

spencerchubb

1,275 karma · joined December 3, 2021

https://twitter.com/spencer_chubb
submissionscomments
spencerchubb··on Where are the builders?
So it sounds like you agree there is not a lack of new trends :)

There are plenty of new ideas and trends. However most people do not pursue them because we're too young, need job stability, etc

spencerchubb··on Where are the builders?
This is a common misconception. With hindsight we can say "It was so easy to succeed back then!"

Why didn't you do those things? Why aren't you trying to catch the next trend? It's extremely difficult to know which trend will actually stick around for the long term.

TikTok could be a current example since they are pushing the shop so hard.

spencerchubb··on Show HN: Pls Fix – Hire big tech employees to appeal account suspensions
On one hand, if you're willing to pay a large sum then it indicates you value your account highly. But on the other hand, this creates perverse incentives for tech employees, and could get them fired.

What if the money was donated instead?

spencerchubb··on Show HN: A VS Code extension to check incompatible CSS
This looks neat! If browsers change, would the extension automatically update or would we have to update the extension or something?
spencerchubb··on Llama3 implemented from scratch
so more people comment on the hn post and it will rank higher in the algo

such as your comment and my comment!

spencerchubb··on Big Companies vs. Startups (2015)
My favorite idea from this post is that you can have a big impact on the company at a small startup, but you can have a bigger impact on the world at a big company. Moving the needle a little bit has a big impact
spencerchubb··on Discord has been using ML to determine the gender and age of some of its users
I would be more surprised if a social app wasn't doing this
spencerchubb··on Bollards: Why and What
I've never thought about in my life. Now that I have read about it, I'm probably going to notice them everywhere.

It seems like a remarkably simple technology that saves lives.

spencerchubb··on Ask HN: Who wants to be hired? (May 2024)
Location: Middle of nowhere Indiana

Remote: No preference

Willing to relocate: Yes

Technologies: LLMs, React, PHP, js/ts, Linux, AWS

Email: spencerchubb at gmail dot com

I am a senior in college graduating in a few weeks. I am the developer of cubingapp.com which is a website for rubiks cubers with about 3k MAU. I am also doing research involving auctions and neural networks.

More details: https://spencerchubb.com

spencerchubb··on Racket Language
Lisp variants only really have linked lists. So yes you can do recursion over them, but for someone who has never done a C-like language before, the arrays and for loops will take getting used to
spencerchubb··on I keep a WTF notebook (2021)
What about the tone bothers you? I've never read How to win friends and influence people. I thought the article is pretty well-written
spencerchubb··on Show HN: Speeding up LLM inference 2x times (possibly)
At first glance, that sounds like it could work. From what I've read, there seems to be two main ways to regain some quality with quantization: post-training which happens after, and quantization-aware training which quantizes during training but leaves the activations and gradients full precision
spencerchubb··on Reddit is taking over Google
I personally love using site:reddit.com when I'm searching on google. I also find myself using tiktok to search for stuff more
spencerchubb··on Reddit is taking over Google
I've said this before and I'll say it again. As companies realize how much people like reddit as a trusted source of information, they will learn how to optimize reddit. Then it will slowly degrade just like google has.
spencerchubb··on Show HN: Speeding up LLM inference 2x times (possibly)
I love this line in the gpu implementation section.

"Readers fresh to GPU programming may ask now - how does it work?

Readers experienced with GPU programming may ask - how the hell does it work?"

spencerchubb··on We are moving to General Availability
capital letters are atypical in many internet circles, particularly younger. so maybe it's a habit from frequenting those circles or a deliberate branding choice
spencerchubb··on Ask HN: Is RAG the Future of LLMs?
When you describe the overlay layer, that sounds similar to the idea of low rank adaptation (LoRA). LoRA is kind of like finetuning, but it doesn't update every parameter, it adds a relatively small number of parameters and finetunes those

Am I understanding what you're describing about the VMs and containers analogy?

spencerchubb··on Ask HN: Is RAG the Future of LLMs?
Good point! If datasets get that large that is an argument in favor of RAG
spencerchubb··on Ask HN: Is RAG the Future of LLMs?
I believe RAG is a temporary hack until we figure out virtually infinite context.

I think LLM context is going to be like cache levels. The first level is small but super fast (like working memory). The next level is larger but slower, and so on.

RAG is basically a bad version of attention mechanisms. RAG is used to focus your attention on relevant documents. The problem is that RAG systems are not trained to minimize loss, it is just a similarity score.

Obligatory note that I could be wrong and it's just my armchair opinion

spencerchubb··on The French aristocrat who understood evolution 100 years before Darwin
This may sound very "hindsight is always 20-20" but why was evolution not always obvious? Traits that help reproduction will increase. Traits that hinder reproduction will decrease. Why was it controversial, and what did people believe before that?
spencerchubb··on More Agents Is All You Need: LLMs performance scales with the number of agents
My understanding is that a meme spreads through memetics similar to how a gene spreads through genetics. Ideas will spread that are fit for reproduction, i.e. communication.

Oftentimes memes are powerful at reproducing, but in science we don't want ideas that are likely to spread. We want ideas that are truthful

spencerchubb··on More Agents Is All You Need: LLMs performance scales with the number of agents
I feel I used the word 'meme' appropriately

"A meme is an idea, behavior, or style that spreads by means of imitation from person to person within a culture and often carries symbolic meaning representing a particular phenomenon or theme."

Paper authors use "All You Need" to allude to the well-known transformer paper, even if their proposed technique is not in fact all you need.

spencerchubb··on More Agents Is All You Need: LLMs performance scales with the number of agents
This trend of "All You Need" in paper titles needs to die. The original "Attention is All You Need" used that title because it is literally true in their case. So many papers just use it as a meme now, and it distracts from the true insight of the paper.
spencerchubb··on Optimize CSS with SAT Solving
I would like a tool similar to this, but for refactoring in a human-readable way instead of minifying. Some tool to automatically look for duplicated css code. In theory, it seems easier for css than for actual programming languages given the simplicity
spencerchubb··on Fuzzing Ladybird with tools from Google Project Zero
Why couldn't the fuzzer be used to discover the bug in the popular browsers?
spencerchubb··on Launch HN: Meticulate (YC W24) – LLM pipelines for business research
Your attention to spelling further shows your meticulousness
spencerchubb··on Building Meta's GenAI infrastructure
GenAI infra is the same as regular AI infra. They used GenAI in the title because it's a buzzword.
spencerchubb··on Building Meta's GenAI infrastructure
All this compute and my Instagram Reels feed still isn't as good as my TikTok feed
spencerchubb··on Yi: Open Foundation Models by 01.AI
An AI model takes a specific input and gives a specific output. In the case of a language model, it takes in text and outputs text. They can not just execute arbitary code

When you download from huggingface, you are downloading the parameters of the model. The parameters are basically just a bunch of numbers, and they had to find the parameters by training the model

spencerchubb··on Tesla starts shipping $3k Cybertruck tent, looks nothing like what was unveiled
You should be more positive. You do get to claim partial credit for the teams you were on :)

If you were very early to the team, you get to claim a larger fraction of credit as well

← PreviousPage 9 of 16Next →