HNHacker News
TopNewBestAskShowJobs

spencerchubb

1,275 karma · joined December 3, 2021

https://twitter.com/spencer_chubb
submissionscomments
spencerchubb··on Netflix buffering issues: Boxing fans complain about Jake Paul vs. Mike Tyson
I would assume not because twitch runs on aws. I think netflix engineers haven't optimized as much for livestreaming like twitch has
spencerchubb··on Netflix buffering issues: Boxing fans complain about Jake Paul vs. Mike Tyson
livestream is quite different from streaming pre-processed video, so I'm not surprised by the scaling issues
spencerchubb··on The online sports gambling experiment
What about the negative externalities? Sports betting has led to more mental illness, bankruptcy, and people who resort to government benefits and crime. The libertarian philosophy is generally live and let live as long as it doesn't harm others.
spencerchubb··on The death and life of prediction markets at Google
There are some ways to mitigate this. You can make the prize pool big enough that people will be motivated, but small enough that people won't try to sway the outcome. Also, prediction markets can be very useful for making predictions about competitors, because it would be rather difficult to influence what happens at a competing company.
spencerchubb··on Chegg is on its last legs after ChatGPT sent its stock down
You gotta look even earlier than 2023. Their net income peaked in 2021 at 776 million for the year, growing 20% year over year. That's a 20x valuation, which is honestly not that crazy considering the bubble of 2021.
spencerchubb··on Chegg is on its last legs after ChatGPT sent its stock down
This is just an anecdote, but I'm seeing more ads than ever for chegg, and I'm not in school. Wondering if it's coincidence or a last ditch effort to boost metrics.
spencerchubb··on Can humans say the largest prime number before we find the next one?
Underrated reason for tiktok's success: shorter videos lend themselves way better to recommendation algorithms, because you have better data about what users want to see
spencerchubb··on The deep learning boom caught almost everyone by surprise
in insurance we use older statistical methods that are easily interpretable, because we are required to have rates approved by departments of insurance
spencerchubb··on The deep learning boom caught almost everyone by surprise
> vertebrates across the board require less data to learn new things

the human brain is absolutely inundated with data, especially from visual, audio, and kinesthetic mediums. the data is a very different form than what one would use to train a CNN or LLM, but it is undoubtedly data. newborns start out literally being unable to see, and they have to develop those neural pathways by taking in the "pixels" of the world for every millisecond of every day

spencerchubb··on An embarrassingly simple approach to recover unlearned knowledge for LLMs
Correct me if I'm wrong, but isn't there no effect on a floating point operation if you make the numbers more precise?
spencerchubb··on I couldn't find a free, no-login, no-AI checklist app–so I built one
What are your needs that can't be satisfied by default notes app on phones?
spencerchubb··on Get me out of data hell
When someone leaves have you ever had code dumped on you that you've never read before and has no tests? If you don't even know what the code is supposed to do, you can't write tests for it.

The way I handled it is: when people come to me with a bug, then I fix it and add a test for it.

spencerchubb··on Get me out of data hell
When LLMs can do the following, they might be able to fix data hell:

- Negotiate with different teams to figure out what a field means

- Be told that a field should be converted from one format to another, but oh wait it's causing errors somewhere downstream because it was told the wrong instructions

- People come to you with some issue about the code you maintain, and you dig enough to realize the root cause is another team's code

spencerchubb··on Get me out of data hell
> my friends stuck writing 195 step DAGs to transform log files from s3 into s3 so they can eventually land in s3

sounds similar to what I do! on my team, it does actually serve a purpose, it's just hard to see the forest from the trees sometimes

spencerchubb··on Show HN: A minimalist (brutalist?) website for sharing all your links
Sorry I should clarify. I didn't mean winning in terms of the website's success, I meant in terms of finding sites that can work without javascript
spencerchubb··on Show HN: A minimalist (brutalist?) website for sharing all your links
I'm curious, how many websites can you even use without javascript? I imagine it would not be many, and would be a losing battle.
spencerchubb··on 34x34x34 Rubik's Cube
I love the Roux method! I just went to a competition this weekend and got my personal record of 9.39 second average with Roux.

The unfortunate part is that beginner tutorials for Roux kind of suck.

spencerchubb··on Real GDP (purchasing power parity)
It's a lot more fun to picture the cia as people wearing disguises to infiltrate foreign governments
spencerchubb··on Moonshine, the new state of the art for speech to text
So they're claiming SOTA because they compare against OpenAI as SOTA. What about Groq or fal.ai?
spencerchubb··on Writes and Write-Nots
Devolved by what metric?
spencerchubb··on Welcome to the Era of the $20k Family Car Insurance Bill
The family has 6 drivers and 6 vehicles, they live in california, and don't have the best driving record. That's $278 per person per month, so really not that outrageous
spencerchubb··on Show HN: Erin – Open-source and self-hosted TikTok feed for your own videos
Do you really think it is semi manual? I guess I never considered the possibility, but I would be surprised. It seems like there are way too many niches and sub-niches, and the algorithm works rather quickly when news happens
spencerchubb··on What docs-as-code means
This article mostly seems to be about documentation for users of the code, not documentation within the code.
spencerchubb··on Accountability sinks
It's much easier to hold an algorithm accountable than an organization of humans. You can reprogram an algorithm. But good look influencing an organization to change
spencerchubb··on Crokinole
What advantage could one get by not having any cheeks in contact with the chair?
spencerchubb··on Reflections on Palantir
That's Peter Thiel's normal demeanor. He stutters and talks awkwardly about every topic
spencerchubb··on How I Experience Web Today (2021)
When I tried to leave the page it said "Changes you made may not be saved" That was a nice touch
spencerchubb··on Stop Ignoring Your High Performers
Is it the case that high performers are bored by it? Or companies don't reward it with promotions?
spencerchubb··on The 1/8th Sleep
Not just venture capital. Established businesses too

Investors like recurring revenue because it's simple to understand. You can just slap a multiple on it to figure out the valuation of a business

spencerchubb··on FLUX is fast and it's open source
That's essentially what diffusion does, except it doesn't have clear boundaries between "scene graph" and "full image". It starts out noisy and adds more detail gradually
← PreviousPage 2 of 16Next →