jin

software engineer @microsoft

[jin @_jinyeom]

#개발자_트친소

반가워요! 시애틀에서 머신러닝 응용쪽으로 일하고 있습니다. 최근엔 배우고 싶은게 많아서 AI외의 주제들도 많이 팔로우/리트윗하고 있어요. 잘 지내봐요 ☺️

[jin @_jinyeom]

@InternetH0F Bro that’s just debt

[Yann LeCun @ylecun]

I-JEPA: Efficient method for Self-Supervised Learning of image features.
No need for data augmentation, just masking.
Joint embedding predictive architecture, not generative.
And it's open source, of course.

Blog: https://t.co/ZuouZgeEMC
Paper: https://t.co/BoHSnELyw8
Code & models: https://t.co/DgS9XiwnMz

[jin @_jinyeom]

@ChristophAdami (When I say most EAs, I really mean the most popular kinds in AI right now, which are typically some variant of ES)

[jin @_jinyeom]

@ChristophAdami I think this is interesting because most EAs don’t evaluate complexity (other than GP and NEAT). Maybe this is what’s missing in progress in EC.

[Joel Lehman @joelbot3000]

"Machine Love" - new paper w/ LM expts: Leaving to humans what is human (emotions/relationships/autonomy), is there any conception of love fitting for machines to embody? One idea: to provide unconditional support for humans to pursue their own growth.
https://t.co/cVl6bB02TF

[jin @_jinyeom]

@deliprao We’re burning old oil to get old electricity to burn new oil to get new electricity

[Luiz Pessoa @PessoaBrain]

𝗥𝗲𝘄𝗮𝗿𝗱 𝗶𝘀 𝗻𝗼𝘁 𝗡𝗲𝗰𝗲𝘀𝘀𝗮𝗿𝘆: How to Create a
Compositional Self-Preserving Agent
for Life-Long Learning

Absolutely tantalizing by @no_reward_for_u
https://t.co/IHoOX089c5

[flavio @flaviocopes]

Web3 devs turning into AI devs

[Viacheslav Biriukov @brk0v]

mmap() is a tremendously important Linux syscall. Many open-source and in-house databases and KV storages use it internally to simplify data access. That's why it's so essential for SREs to understand its interaction with Page Cache: https://t.co/SaLhniuPpK

[jin @_jinyeom]

Arch @OsEndeavour

[jin @_jinyeom]

@evolvingstuff We’ve already seen so many things that will supposedly “make SWE trivial”; none of them have replaced engineers 🤷‍♂️

[jin @_jinyeom]

I guess I’m mostly talking about LLM-related projects. Man, am I tired of hearing about those.

[jin @_jinyeom]

I feel like whatever Google is trying to do with AI may be almost as ludicrous as what Meta is doing with metaverse; yet the public’s perception is seemingly different. I’m starting to wonder if pursuing VR makes more sense than AI at this point.

[jin @_jinyeom]

This is all of course from a perspective of engineering, not science. AI might always be interesting as science projects, but how long until the public gets tired of all the PR research projects that cost more to execute than the profit?

[jin @_jinyeom]

@davecraige @deliprao I wouldn’t group you guys with the experts lmao let the experts do what they’re good at

[jin @_jinyeom]

@noahconst @savvyRL @dhgarrette @Chitwan_Saharia @wchan212 @ada_rob @sharan0909 @mo_norouzi @irinablok Huh I’ll be damned 🫢 https://t.co/dV5fnEsJnl

[jin @_jinyeom]

@patrickmineault Any chance this is related to the language encoder’s tokenizer?

[jin @_jinyeom]

@cesarberardini @benedictevans I wouldn’t say they’re the same — people already carried phones; smart phones got better without changing the form factor too much. Everyday people don’t wear headsets in 2022, they’re not gonna suddenly start in 2023.

[jin @_jinyeom]

@scrumtuous Building a website engine