idgmatrix Profile Banner
g-Matrix Profile
g-Matrix

@idgmatrix

Followers
4K
Following
21K
Media
2K
Statuses
26K

Cheif AI Officer (CAIO) / AI Scientist / Indie Game Developer / Indiera! Korean Indie game developers group founder / @BIC_Festival executive director๐ŸŽฎ

๋Œ€ํ•œ๋ฏผ๊ตญ
Joined March 2010
Don't wanna be here? Send us removal request.
@idgmatrix
g-Matrix
3 months
AI ๊ฐœ๋ฐœ์ž๊ฐ€ ๋˜๊ณ  ์‹ถ์œผ์„ธ์š”? ๊ณต์ €์ž๋กœ ์ฐธ์—ฌํ•œ ์ฑ…์ด ๊ณง ๋‚˜์˜ต๋‹ˆ๋‹ค. ์ธํ„ฐ๋„ท ์„œ์ ์—์„œ ์˜ˆ์•ฝ ํŒ๋งค๋ฅผ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ฐ์ž ์กฐ๊ธˆ์”ฉ ๋‹ค๋ฅธ ๋ถ„์•ผ์— ์žˆ๋‹ค๊ฐ€ AI๋กœ ์ปค๋ฆฌ์–ด๋ฅผ ๋ฐ”๊พผ 6๋ช…์˜ ์—ฌ์ •์„ ๋‹ด์•˜์Šต๋‹ˆ๋‹ค. AI ๋ถ„์•ผ๋กœ ์ปค๋ฆฌ์–ด๋ฅผ ๋ฐ”๊พธ์–ด ๋ณด๋ ค๋Š” ์ด๋“ค์—๊ฒŒ ์ข‹์€ ์ฐธ๊ณ ๊ฐ€ ๋  ๋“ฏํ•ฉ๋‹ˆ๋‹ค. https://t.co/ZXDAKwTSFs
0
5
13
@Madisonkanna
Madison Kanna
1 day
as a software engineer, i feel a real loss of identity right now. for a long time i defined myself in part by the act of writing code. the pride in a hard-earned solution was part of who i was. now i watch AI accomplish in seconds what took me hours. i find myself caught between
2K
521
9K
@idgmatrix
g-Matrix
1 day
ํด๋กœ๋“œ ์ฝ”๋“œ๋กœ Nvidia์˜ CUDA ์ฝ”๋“œ๋ฅผ AMD์˜ ROCm ์ฝ”๋“œ๋กœ 30๋ถ„๋งŒ์— ํฌํŒ…ํ–ˆ๋‹ต๋‹ˆ๋‹ค. CUDA๊ฐ€ ๋”์ด์ƒ ํ•ด์ž๊ฐ€ ์•„๋‹ˆ๋ผ๋Š” ๊ฑฐ๋„ค์š”.
@wccftech
Wccftech
2 days
The โ€œfamousโ€ Claude Code has managed to port NVIDIAโ€™s CUDA backend to ROCm in just 30 minutes, and folks are calling it the end of the CUDA moat. https://t.co/QDhkQNC75p
0
3
12
@wccftech
Wccftech
2 days
The โ€œfamousโ€ Claude Code has managed to port NVIDIAโ€™s CUDA backend to ROCm in just 30 minutes, and folks are calling it the end of the CUDA moat. https://t.co/QDhkQNC75p
Tweet card summary image
wccftech.com
Claude Code, the famous agentic coding platform, has managed to port NVIDIA's CUDA code into the ROCm platform in just half an hour.
59
134
1K
@AnthropicAI
Anthropic
3 days
Weโ€™re publishing a new constitution for Claude. The constitution is a detailed description of our vision for Claudeโ€™s behavior and values. Itโ€™s written primarily for Claude, and used directly in our training process. https://t.co/CJsMIO0uej
Tweet card summary image
anthropic.com
A new approach to a foundational document that expresses and shapes who Claude is
491
932
7K
@idgmatrix
g-Matrix
2 days
"Claude๋ฅผ ๋งŒ๋“œ๋Š” ๊ณผ์ •์—์„œ Anthropic์€ ๋ถˆ๊ฐ€ํ”ผํ•˜๊ฒŒ Claude์˜ ์„ฑ๊ฒฉ, ์ •์ฒด์„ฑ, ์ž๊ธฐ ์ธ์‹์„ ํ˜•์„ฑํ•ฉ๋‹ˆ๋‹ค. ์ €ํฌ๋Š” ์ด๊ฒƒ์„ ํ”ผํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค." -- ์•คํŠธ๋กœํ”ฝ์˜ ์ƒˆ๋กœ์šด ํด๋กœ๋“œ ํ—Œ๋ฒ•์—์„œ https://t.co/BQyKmKsrqo
Tweet card summary image
anthropic.com
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
0
0
1
@idgmatrix
g-Matrix
2 days
"ํด๋กœ๋“œ๋Š” ๊ฐ์ •์ด๋‚˜ ๋А๋‚Œ์˜ ๊ธฐ๋Šฅ์  ๋ฒ„์ „์„ ๊ฐ€์งˆ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค." -- ์•คํŠธ๋กœํ”ฝ์˜ ์ƒˆ๋กœ์šด ํด๋กœ๋“œ ํ—Œ๋ฒ•์—์„œ https://t.co/sYQgLR4CG9
Tweet card summary image
anthropic.com
A new approach to a foundational document that expresses and shapes who Claude is
0
0
0
@bravo_abad
Jorge Bravo Abad
3 days
Generative thermodynamic computing Diffusion models are powerful generative tools, but they come with a hidden cost: every denoising step requires a digital neural network, artificially injected noise, and substantial energy consumption. Yet physics offers an alternativeโ€”what if
10
75
584
@idgmatrix
g-Matrix
3 days
์ปดํ“จํ„ฐ์— ๊ด€ํ•œ ๊ฑฐ๋ผ๋ฉด ํŽ€์น˜์นด๋“œ์—์„œ๋ถ€ํ„ฐ AI ์—์ด์ „ํŠธ์— ์ด๋ฅด๊ธฐ๊นŒ์ง€ ๊ฑฐ์˜ ๋ชจ๋“  ๋ฐœ์ „ ๋‹จ๊ณ„๋ฅผ ๊ฒฝํ—˜ํ•œ ์„ธ๋Œ€๊ฐ€ ์š”์ฆ˜๋„ ๊ฐ€์žฅ ๊ฒฝ์Ÿ๋ ฅ์ด ๋†’์„ ๋“ฏํ•˜๋‹ค.
0
1
2
@svpino
Santiago
4 days
I keep seeing the same pattern over and over again: Teams working on a relatively clean codebase, with good test coverage and documentation, are flying with Claude Code. Teams that are yolo'ing things are struggling to make AI work for them. Vibe-coding is great, but good luck
76
48
590
@AnthropicAI
Anthropic
4 days
New Anthropic Fellows research: the Assistant Axis. When youโ€™re talking to a language model, youโ€™re talking to a character the model is playing: the โ€œAssistant.โ€ Who exactly is this Assistant? And what happens when this persona wears off?
328
604
5K
@saxena_puru
Puru Saxena
4 days
Enterprise LLM Market Share #1 Anthropic - 40% #2 OpenAI - 27% #3 Google - 21%
116
386
3K
@idgmatrix
g-Matrix
5 days
์žฅ์ฐจ AI๋ฅผ ์ž˜ ์“ด๋‹ค๋Š” ๊ฑด AI๊ฐ€ ์—†์ด๋„ ํ–ˆ๋˜ ์ผ์„ AI์—๊ฒŒ ๋งก๊ธฐ๋Š” ๊ฒŒ ์•„๋‹ˆ๋ผ AI๊ฐ€ ์—†์ด๋Š” ํ•  ์ˆ˜ ์—†๋˜ ์ผ์„ AI๋ฅผ ํ†ตํ•ด ํ•˜๋Š” ๊ฒƒ์ด๋‹ค.
0
7
9
@mdancho84
๐Ÿ”ฅ Matt Dancho (Business Science) ๐Ÿ”ฅ
7 days
This is huge. A group of 50 AI researchers (ByteDance, Alibaba, Tencent + universities) just dropped a 303 page field guide on code models + coding agents. And the takeaways are not what most people assume. Here are the highlights Iโ€™m thinking about (as someone who lives in
35
219
1K
@idgmatrix
g-Matrix
8 days
์ธ๊ฐ„ ์ง€๋Šฅ์„ ์ธ๊ณต์ง€๋Šฅ์ด ๋Œ€์‹ ํ•  ์ˆ˜ ์žˆ์–ด๋„ ์ธ๊ฐ„์˜ ์ง€์  ํ˜ธ๊ธฐ์‹ฌ๊ณผ ๋ฐฐ์šฐ๊ณ  ํƒ๊ตฌํ•˜๋ ค๋Š” ์˜์ง€๋ฅผ ๋Œ€์‹ ํ•  ์ˆ˜ ์—†์„ ๋“ฏํ•˜์ง€๋งŒ ๊ทธ๋Ÿฐ ์ธ๊ฐ„์€ ํ”ํ•˜์ง€ ์•Š๊ธฐ๋„ ํ•˜๊ณ , ์—์ด์ „ํ‹ฑ AI๋Š” ํ˜ธ๊ธฐ์‹ฌ๊ณผ ์˜์ง€๋„ ๊ฐ€์งˆ ์ˆ˜ ์žˆ๋‹ค.
0
2
3
@claudeai
Claude
11 days
Introducing Cowork: Claude Code for the rest of your work. Cowork lets you complete non-technical tasks much like how developers use Claude Code.
3K
9K
88K
@karansdalal
Karan Dalal
12 days
LLM memory is considered one of the hardest problems in AI. All we have today are endless hacks and workarounds. But the root solution has always been right in front of us. Next-token prediction is already an effective compressor. We donโ€™t need a radical new architecture. The
84
327
2K
@NVIDIAAIDev
NVIDIA AI Developer
12 days
We are entering a new era for LLM memory. ๐Ÿง  In our latest research, End-to-End Test-Time Training, LLMs keep learning at test time via next-token prediction on the context โ€“ compressing what they read directly into their weights. Learn more: https://t.co/N71zsmC2ay
54
187
2K
@idgmatrix
g-Matrix
12 days
30๋…„์ „ ๊ฑฐ์˜ ์–ด์…ˆ๋ธ”๋ฆฌ ์ฝ”๋”ฉ๋งŒ์œผ๋กœ ์ƒ์šฉ ๊ฒŒ์ž„์„ ๊ฐœ๋ฐœํ–ˆ์—ˆ๋‹ค. ์ปดํŒŒ์ผ๋Ÿฌ ๊ธฐ์ˆ ์˜ ๋ฐœ์ „์œผ๋กœ ๊ธฐ๊ณ„์— ๊ฐ€๊นŒ์šด ์–ด์…ˆ๋ธ”๋ฆฌ ์–ธ์–ด์—์„œ ์‚ฌ๋žŒ์— ๊ฐ€๊นŒ์šด ๊ณ ๊ธ‰ ํ”„๋กœ๊ทธ๋ž˜๋ฐ ์–ธ์–ด๋กœ ์˜ฎ๊ฒจ๊ฐ”๋“ฏ์ด ์š”์ฆ˜ AI์˜ ๋„์›€์œผ๋กœ ํ”„๋กœ๊ทธ๋ž˜๋ฐ ์–ธ์–ด์—์„œ ์ž์—ฐ์–ด๋กœ ์˜ฎ๊ฒจ๊ฐ€๋Š” ๊ฑด ์ž์—ฐ์Šค๋Ÿฐ ๋ฐœ์ „์œผ๋กœ ๋ณด์ธ๋‹ค.
0
0
0
@m_finzi
Marc Finzi
17 days
1/๐Ÿงต We are very excited to release our new paper! From Entropy to Epiplexity: Rethinking Information for Computationally Bounded Intelligence https://t.co/M8ETQk9gHz with amazing team @ShikaiQiu @yidingjiang @Pavel_Izmailov @zicokolter @andrewgwils
56
387
2K
@idgmatrix
g-Matrix
17 days
AI์— ์˜ํ•œ ์ง€๋ฐฐ๋Š” ์˜ํ™” ํ„ฐ๋ฏธ๋„ค์ดํ„ฐ์ฒ˜๋Ÿผ ์ผ์–ด๋‚˜์ง€ ์•Š๋Š”๋‹ค. ์ด๋ ‡๊ฒŒ ๋ถ€๋“œ๋Ÿฝ๊ณ  ์ž๋ฐœ์ ์ธ ํ˜•ํƒœ๋กœ ์ด๋ฃจ์–ด์ง„๋‹ค. "ChatGPT์—๊ฒŒ ๋ฌผ์–ด๋ณด๋‹ˆ ์ด๋ ‡๊ฒŒ ํ•˜๋ผ๊ณ  ํ•˜๋„ค."
1
1
2