Continuous Learning_Startup & Investment
2.43K subscribers
513 photos
5 videos
16 files
2.74K links
We journey together through the captivating realms of entrepreneurship, investment, life, and technology. This is my chronicle of exploration, where I capture and share the lessons that shape our world. Join us and let's never stop learning!
Download Telegram
https://www.instagram.com/p/CpcLIYpPsAY/?igshid=MjkzY2Y1YTY%3D

Hospitals face challenges when negotiating with insurance companies. Customers' problems present opportunities for businesses.

Both hospitals and individuals, including myself, face challenges in collecting necessary documents, applying for insurance, and negotiating with insurance companies.

Could we address this problem?
Google is investing in Runway, a New Yorkโ€“based startup that lets customers generate video from text descriptions using artificial intelligence it pioneered, at a valuation of around $1.5 billion including the new capital, according to two people familiar with the matter.

The investment, part of a roughly $100 million funding round, underscores the fierce competition among cloud providers to get close to companies with cutting-edge AI services that could become major cloud customers or acquisition targets in the future. Amazon Web Services has touted Runway, which generates relatively little revenue from its video-editing tools, as a key AI-startup customer but Runway is now expected to rent cloud servers from Google, said one person briefed about the deal.

https://www.theinformation.com/articles/google-invests-in-ai-startup-runway-to-wrest-cloud-business-from-aws?utm_campaign=article_email&utm_content=article-10624&utm_source=sg&utm_medium=email&rc=ocojsj
AI ๋Œ€ํ•ญํ•ด์‹œ๋Œ€๋„ค์š” ใ…Žใ…Ž ์„ค๋ ˆ๋„ค์š” ๐Ÿ™‚
๐Ÿ‘1
ํ™”์žฅํ’ˆ, ์ „์ž์ œํ’ˆ, ์ž๋™์ฐจ, ํŒจ์…˜&์„ฌ์œ , ์‹ํ’ˆ, B2B ๋ถ„์•ผ์—์„œ ํ•ด์™ธ ๋งˆ์ผ€ํŒ…(์˜๋ฏธ๊ถŒ)์„ ๋‹ด๋‹นํ•˜์‹œ๋Š” ๋ถ„ ํ˜น์€ ์ง€์ธ์ด ์žˆ์œผ์‹ค๊นŒ์š”?
๊ด€์‹ฌ์žˆ์œผ์‹ค๋งŒํ•œ ์ฃผ์ œ๋ฅผ ์ œ์•ˆ๋“œ๋ฆฌ๋ ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค!
@MatthewMinseokKim ์œผ๋กœ ์—ฐ๋ฝ์ฃผ์‹œ๋ฉด ๊ฐ์‚ฌํ•˜๊ฒ ์Šต๋‹ˆ๋‹ค ๐Ÿ™
Continuous Learning_Startup & Investment
ํ™”์žฅํ’ˆ, ์ „์ž์ œํ’ˆ, ์ž๋™์ฐจ, ํŒจ์…˜&์„ฌ์œ , ์‹ํ’ˆ, B2B ๋ถ„์•ผ์—์„œ ํ•ด์™ธ ๋งˆ์ผ€ํŒ…(์˜๋ฏธ๊ถŒ)์„ ๋‹ด๋‹นํ•˜์‹œ๋Š” ๋ถ„ ํ˜น์€ ์ง€์ธ์ด ์žˆ์œผ์‹ค๊นŒ์š”? ๊ด€์‹ฌ์žˆ์œผ์‹ค๋งŒํ•œ ์ฃผ์ œ๋ฅผ ์ œ์•ˆ๋“œ๋ฆฌ๋ ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค! @MatthewMinseokKim ์œผ๋กœ ์—ฐ๋ฝ์ฃผ์‹œ๋ฉด ๊ฐ์‚ฌํ•˜๊ฒ ์Šต๋‹ˆ๋‹ค ๐Ÿ™
๐Ÿš€ ์ฑ—๋ด‡ ์—”์ง„ ์ตœ์ ํ™”(Chatbot Engine Optimization)์˜ ์‹œ๋Œ€๊ฐ€ ์™”์Šต๋‹ˆ๋‹ค. ์ค€๋น„๋˜์…จ๋‚˜์š”? ๐Ÿš€
ChatGPT๋Š” ํ˜„์žฌ ํ•˜๋ฃจ ํ™œ์„ฑ ์‚ฌ์šฉ์ž ๋ฐฉ๋ฌธ์ด 6000๋งŒ, ์›”๊ฐ„ ์‚ฌ์šฉ์ž ๋ฐฉ๋ฌธ์€ ๋ฌด๋ ค 18์–ต์— ์ด๋ฆ…๋‹ˆ๋‹ค. Yahoo, Google, Facebook, YouTube, TikTok ๋“ฑ์˜ ์‚ฌ๋ก€๋ฅผ ๋ดค์„ ๋•Œ, ์‚ฌ์šฉ์ž๋“ค์ด ๋งŽ์€ ์‹œ๊ฐ„์„ ๋ณด๋‚ด๋Š” ์„œ๋น„์Šค๋Š” ์ƒˆ๋กœ์šด ๊ด‘๊ณ  ํ”Œ๋žซํผ์œผ๋กœ ์„ฑ์žฅํ•ด์™”์Šต๋‹ˆ๋‹ค. ๋งˆ์ฐฌ๊ฐ€์ง€๋กœ ์•ž์œผ๋กœ ๊ฐ€์žฅ ํฐ ๊ด‘๊ณ  ํ”Œ๋žซํผ์€ ChatGPT์™€ ๊ฐ™์€ AI ์ฑ—๋ด‡๋“ค์ด ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
๊ณผ๊ฑฐ์— Facebook, Youtube, TikTok์ด ์ดˆ๊ธฐ์— ๋“ฑ์žฅํ–ˆ์„ ๋•Œ ๋งˆ์ผ€ํŒ…์„ ํšจ๊ณผ์ ์œผ๋กœ ์ง„ํ–‰ํ•œ ๊ธฐ์—…๋“ค์ด ์ƒˆ๋กœ์šด ์„ฑ์žฅ์˜ ๊ธฐํšŒ๋ฅผ ์žก์•˜๋˜ ๊ฒƒ์ฒ˜๋Ÿผ, AI Chatbot์„ ํ†ตํ•ด ์ž์‹ ์˜ ์ œํ’ˆ ์ •๋ณด๋ฅผ ํšจ๊ณผ์ ์œผ๋กœ ์ „๋‹ฌํ•˜๋Š” ๊ฒƒ์€ ๋ชจ๋“  ๊ธฐ์—…์—๊ฒŒ ์ค‘์š”ํ•œ ๊ธฐํšŒ๊ฐ€ ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
์ด๋ฏธ Bay Area์˜ ๋ฒ•๋ฅ ์‚ฌ๋ฌด์†Œ์™€ ์ „์ž์ƒ๊ฑฐ๋ž˜ ํšŒ์‚ฌ๋“ค ์ค‘ ChatGPT๋ฅผ ํ†ตํ•ด ์‹  ๊ทœ ๊ณ ๊ฐ์„ ํ™•๋ณดํ•œ ๊ฒฝ์šฐ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค. AI ์ฑ—๋ด‡ ์‹œ์žฅ์€์ด๋ฏธ ๋น ๋ฅด๊ฒŒ ์„ฑ์žฅํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด๋Š” ๋””์ง€ํ„ธ ๋งˆ์ผ€ํŒ…์—์„œ SEO๋งŒํผ ์ค‘์š”ํ•œ ์—ญํ• ์„ ํ•˜๊ฒŒ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
SEO๊ฐ€ ์ด๋ฏธ ํฌํ™”๋œ ๊ฒฝ์Ÿ ์‹œ์žฅ์ธ ๋ฐ˜๋ฉด, ์ฑ—๋ด‡ ์—”์ง„ ์ตœ์ ํ™”(ChatEO)์˜ ์‹œ๋Œ€๋Š” ๋ง‰๋ฐ”์ง€๋ฅผ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ChatEO๋Š” AI๊ฐ€ ํŠน์ • ์ œํ’ˆ์˜ ์ •๋ณด๋ฅผ ์ž˜ ํ•™์Šตํ•˜๊ณ  ๊ฒ€์ƒ‰ํ•  ์ˆ˜ ์žˆ๋„๋ก ์ •๋ณด๋ฅผ ์ž˜ ์ •๋น„ํ•˜๊ณ , ํŠน์ • ์ œํ’ˆ์„ ์†Œ๋น„ํ•˜๋Š” ๊ณ ๊ฐ์—๊ฒŒ ๋งž๋Š” AI ๊ธฐ๋ฐ˜ ์„œ๋น„์Šค๋ฅผ ์ค€๋น„ํ•˜๋Š” ๊ฒƒ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค.
๋””์ง€ํ„ธ ๋งˆ์ผ€ํŒ…์— ์žˆ์–ด์„œ ํฅ๋ฏธ์ง„์ง„ํ•œ ์‹œ๊ธฐ์ž…๋‹ˆ๋‹ค. AI ์ฑ—๋ด‡๊ณผ ChatEO์˜ ๋ถ€์ƒ์€ ๊ณ ๊ฐ ์ฐธ์—ฌ์™€ ๋””์ง€ํ„ธ ๊ด‘๊ณ ์˜ ์ง€ํ˜•๋„๋ฅผ ํ˜์‹ ์ ์œผ๋กœ ๋ฐ”๊ฟ€ ์ˆ˜ ์žˆ์„ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด ๋ณ€ํ™”๋ฅผ ํšจ๊ณผ์ ์œผ๋กœ ๋‹ค๋ฃฐ ์ˆ˜ ์žˆ๋Š” ๊ธฐ์—…๋“ค์€ ์ด ์ƒˆ๋กœ์šด ํ™˜๊ฒฝ์—์„œ ๋‹จ์ˆœํžˆ ์ƒ์กดํ•˜๋Š” ๊ฒƒ์„ ๋„˜์–ด ๋ฒˆ์˜ํ•  ์ˆ˜ ์žˆ์„ ๊ฒƒ ์ž…๋‹ˆ๋‹ค.
ํ•œ๊ตญ์ด ๊ธ€๋กœ๋ฒŒํ•˜๊ฒŒ ์ˆ˜์ถœํ•˜๋Š” ํ’ˆ๋ชฉ์ธ ํ™”์žฅํ’ˆ, ์ „์ž์ œํ’ˆ, ์ž๋™์ฐจ, ํŒจ์…˜&์„ฌ์œ , ์‹ํ’ˆ, B2B ๋“ฑ์ด๊ณ  ๊ด€๋ จ๋œ ์—…๊ณ„์—์„œ ์˜๋ฏธ๊ถŒ ๋Œ€์ƒ ํ•ด์™ธ ๋งˆ์ผ€ํŒ…์„ ๋‹ด๋‹นํ•˜์‹œ๋Š” ๋ถ„๋“ค์€ ์ด ํŠธ๋ Œ๋“œ๋ฅผ ์ •ํ™•ํžˆ ์ดํ•ดํ•  ํ•„์š”๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.
์ด ์ƒˆ๋กœ์šด ํŠธ๋ Œ๋“œ์— ๊ด€์‹ฌ์ด ์žˆ์œผ์‹  ๋ถ„๋“ค, ํ˜น์€ ๋” ์ž์„ธํ•œ ์ด์•ผ๊ธฐ๋ฅผ ๋‚˜๋ˆ„๊ณ  ์‹ถ์œผ์‹  ๋ถ„๋“ค์€ ์•„๋ž˜ ๋งํฌ๋ฅผ ํด๋ฆญํ•ด์ฃผ์‹œ๊ธฐ ๋ฐ”๋ž๋‹ˆ๋‹ค. ๐Ÿ™
https://forms.gle/vqChSP989extfTLj7
#AI #์ฑ—๋ด‡ #ChatGPT #ChatEO #๋””์ง€ํ„ธ๋งˆ์ผ€ํŒ… #SEO
Continuous Learning_Startup & Investment
ํ™”์žฅํ’ˆ, ์ „์ž์ œํ’ˆ, ์ž๋™์ฐจ, ํŒจ์…˜&์„ฌ์œ , ์‹ํ’ˆ, B2B ๋ถ„์•ผ์—์„œ ํ•ด์™ธ ๋งˆ์ผ€ํŒ…(์˜๋ฏธ๊ถŒ)์„ ๋‹ด๋‹นํ•˜์‹œ๋Š” ๋ถ„ ํ˜น์€ ์ง€์ธ์ด ์žˆ์œผ์‹ค๊นŒ์š”? ๊ด€์‹ฌ์žˆ์œผ์‹ค๋งŒํ•œ ์ฃผ์ œ๋ฅผ ์ œ์•ˆ๋“œ๋ฆฌ๋ ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค! @MatthewMinseokKim ์œผ๋กœ ์—ฐ๋ฝ์ฃผ์‹œ๋ฉด ๊ฐ์‚ฌํ•˜๊ฒ ์Šต๋‹ˆ๋‹ค ๐Ÿ™
What Charlie said.

์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์— ์œ ์˜ํ•˜๋ผ
๋งˆ์ง€๋ง‰์€ โ€˜์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ(Self-serving bias)โ€™์ด๋‹ค. ์‚ฌ๋žŒ๋“ค์€ ์„ ์ฒœ์ ์œผ๋กœ ๋‹ค๋ฅธ ์‚ฌ๋žŒ์˜ ์ด์ต๋ณด๋‹ค๋Š” ์ž๊ธฐ ์ž์‹ ์˜ ์ด์ต์—๋งŒ ์ง‘์ค‘ํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ๋‹ค. ์‹ฌ์ง€์–ด ์ž์‹ ์—๊ฒŒ ์ข‹์€ ์ผ์ด๋ฉด ์‚ฌํšŒ์—๋„ ์ข‹๋‹ค๋Š” ์‚ฌ๊ณ ๋ฐฉ์‹์ด ๋ฌด์˜์‹์— ๊น”๋ ค ์žˆ๊ธฐ ๋•Œ๋ฌธ์— ์ด๊ธฐ์ ์ธ ํ–‰๋™์œผ๋กœ ์ด์–ด์ง€๊ธฐ๋„ ํ•œ๋‹ค.
ํ•˜์ง€๋งŒ, ๋ฉ๊ฑฐ๋Š” ์ด๊ฑด ๋Œ€๋‹จํžˆ ์ž˜๋ชป๋œ ์‚ฌ๊ณ ๋ฐฉ์‹์ด๋ผ๋ฉฐ ํ˜„๋ช…ํ•ด์ง€๊ธฐ ์œ„ํ•ด์„œ๋Š” ์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์—์„œ ๋ฒ—์–ด๋‚˜์•ผ ํ•˜์ง€๋งŒ, ๋Œ€๋‹ค์ˆ˜ ์‚ฌ๋žŒ๋“ค์ด ์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์„ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๋Š” ์‚ฌ์‹ค๋„ ์ธ์ •ํ•ด์•ผ ํ•œ๋‹ค๊ณ  ๋งํ–ˆ๋‹ค. ์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์„ ๋ฒ—์–ด๋‚˜๊ธฐ๋Š” ๋Œ€๋‹จํžˆ ์–ด๋ ต๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค.
๊ทธ๋ž˜์„œ ๋ฉ๊ฑฐ๋Š” ์šฐ๋ฆฌ๊ฐ€ ํ–‰๋™ํ•  ๋•Œ๋Š” ์‚ฌ๋žŒ๋“ค์ด ์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์„ ๊ฐ€์ง€๊ณ  ์žˆ๋‹ค๋Š” ๊ฑธ ์—ผ๋‘์— ๋‘๊ณ  ํ–‰๋™ํ•ด์•ผ ํ•œ๋‹ค๋Š” ์ ์„ ์˜ˆ๋ฆฌํ•˜๊ฒŒ ์ง€์ ํ–ˆ๋‹ค. ์—ฌ๊ธฐ์„œ ๋ฉ๊ฑฐ๊ฐ€ ์˜ˆ๋กœ ๋“  ๊ฑด ๋ฏธ๊ตญ ํˆฌ์ž์€ํ–‰ ์‚ด๋กœ๋จผ ๋ธŒ๋ผ๋”์Šค์˜ ๊ณ ๋ฌธ๋ณ€ํ˜ธ์‚ฌ๋‹ค.
๋‹น์‹œ ์‚ด๋กœ๋จผ ๋ธŒ๋ผ๋”์Šค์˜ ์ตœ๊ณ  ๊ฒฝ์˜์ž(CEO)๊ฐ€ ์ž์‚ฌ ์ฑ„๊ถŒ ํŠธ๋ ˆ์ด๋”๊ฐ€ ๊ตญ์ฑ„ ์ž…์ฐฐ์—์„œ ๊ด€๋ จ ๊ทœ์ •์„ ์–ด๊ฒผ๋‹ค๋Š” ๊ฑธ ์•Œ๊ฒŒ ๋๋‹ค. ๊ณ ๋ฌธ๋ณ€ํ˜ธ์‚ฌ๋Š” CEO์—๊ฒŒ ๋‹น๊ตญ์— ๋ณด๊ณ ํ•  ๋ฒ•์  ์˜๋ฌด๋Š” ์—†์ง€๋งŒ, ๊ทธ๋ž˜๋„ ๋ณด๊ณ ํ•˜๋Š” ๊ฒŒ ๋„๋•์ ์ธ ์˜๋ฌด๋ผ๊ณ  ๋งํ–ˆ๋‹ค. ํ•˜์ง€๋งŒ, ์ด๊ฑด ๋ˆ„๊ตฌ๋„ ์›์น˜ ์•Š๋Š” ์ผ์ด์—ˆ๊ธฐ์— CEO๋Š” ๊ณ„์† ๋ฏธ๋ค˜๊ณ  ๊ฒฐ๊ตญ ์‚ด๋กœ๋จผ ๋ธŒ๋Ÿฌ๋”์Šค์˜ ์ฑ„๊ถŒ ์Šค์บ”๋“ค๋กœ ๋น„ํ™”๋˜๋ฉด์„œ CEO๋ฟ ์•„๋‹ˆ๋ผ ๊ณ ๋ฌธ๋ณ€ํ˜ธ์‚ฌ๊นŒ์ง€ ์ผ์ž๋ฆฌ๋ฅผ ์žƒ์—ˆ๋‹ค.
์ด์— ๋Œ€ํ•ด, ๋ฉ๊ฑฐ๋Š” ๋ฏธ๊ตญ ๊ฑด๊ตญ์˜ ์•„๋ฒ„์ง€ ๋ฒค์ž๋ฏผ ํ”„๋žญํด๋ฆฐ์˜ ๋ง์„ ๋นŒ๋ ค โ€œ๋งŒ์•ฝ ์„ค๋“ํ•˜๋ ค๊ณ  ํ•œ๋‹ค๋ฉด ์ด์„ฑ์ด ์•„๋‹ˆ๋ผ ์ด์ต์— ํ˜ธ์†Œํ•˜๋ผโ€œ๊ณ  ๋งํ–ˆ๋‹ค. ์ฆ‰, CEO์˜ ์ž๊ธฐ ์œ„์ฃผ ํŽธํ–ฅ์„ ๊ณ ๋ คํ•ด์„œ CEO์˜ ์ด์ต์— ํ˜ธ์†Œํ•ด์•ผ ํ–ˆ๋‹ค๋Š” ์˜๋ฏธ๋‹ค.
๋งŒ์•ฝ ๊ณ ๋ฌธ๋ณ€ํ˜ธ์‚ฌ๊ฐ€ CEO์—๊ฒŒ โ€œ์ด ์ผ์„ ๊ทธ๋Œ€๋กœ ๋†”๋‘๋ฉด ๊ฒฐ๊ตญ ์‚ฌ์žฅ๋‹˜์˜ ๊ฒฝ๋ ฅ์„ ๋ง๊ฐ€๋œจ๋ฆฌ๊ณ  ๋ˆ๊ณผ ์ง€์œ„, ๋ช…์˜ˆ๋ฅผ ๋ฌด๋„ˆ๋œจ๋ฆด ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ œ ๊ถŒ๊ณ ๋ฅผ ๋”ฐ๋ฅด๋ฉด ๊ทธ๋Ÿฐ ์‚ฌํƒœ๋ฅผ ๋ฐฉ์ง€ํ•  ์ˆ˜ ์žˆ์„ ๊ฒ๋‹ˆ๋‹คโ€œ๋ผ๊ณ  ๋งํ–ˆ๋‹ค๋ฉด CEO์™€ ๊ณ ๋ฌธ๋ณ€ํ˜ธ์‚ฌ ๋ชจ๋‘ ์•„๋ฌด์ผ ์—†์—ˆ์„ ๊ฒƒ์ด๋‹ค.
๐Ÿ‘3
New world to VR?

Appleโ€™s first foray into mixed reality hardware (that is, a blend of real and virtual worlds) is one of the companyโ€™s worst-kept secrets. The tech giant has been acquiring headset-friendly startups for years, and Tim Cook hasnโ€™t been shy about his interests in augmented and virtual reality. Now, though, a headset appears to be ready for a WWDC debut โ€” Appleโ€™s event logo even hints at a lens.

If the rumors are accurate, Appleโ€™s stand-alone device (possibly called Reality Pro) may be more powerful than the Meta Quest Pro and many other high-end headsets. It might offer a 4K resolution per eye, with at least six cameras providing a view of the outside world as well as full body motion tracking. It could also use an M2 chip that outperforms the mobile processors you typically see in stand-alone headsets. Expectations are that it will require an external battery pack and last for just two hours on a charge, but it would be relatively light and slim.

The true star might be the software, though. The Apple headset will reportedly run a new platform (possibly called xrOS) designed with mixed reality in mind. Rather than rely on controllers, you would use hand gestures, gaze detection and Siri voice commands to navigate the interface. An Apple Watch-style crown would toggle between AR and VR, and you could type in mid-air.

Any news about AI from Apple? ๐Ÿง
ํ”„๋กœ์„ธ์Šค ๊ฐ๋…์„ ํ†ตํ•œ ์ˆ˜ํ•™์  ์ถ”๋ก  ํ–ฅ์ƒ
https://openai.com/research/improving-mathematical-reasoning-with-process-supervision?fbclid=IwAR3ICUhp_ALBMBUMOG_vqe1cqCMpLNTzm0yWkbIP2hbeeKX7OF-11_RCXQo

OpenAI๊ฐ€ ๋˜ ๋‹ค๋ฅธ ๋ŒํŒŒ๊ตฌ๋ฅผ ๋งˆ๋ จํ–ˆ๊ตฐ์š”! ์ตœ์ข… ์ •๋‹ต๋ฟ๋งŒ ์•„๋‹ˆ๋ผ ๊ฐ ์ •๋‹ต ๋‹จ๊ณ„๋งˆ๋‹ค ๋ณด์ƒ์„ ์ œ๊ณตํ•˜๋Š” ์ ‘๊ทผ์œผ๋กœ ์ˆ˜ํ•™ ๋ฌธ์ œ ํ’€์ด์—์„œ ์ƒˆ๋กœ์šด ๊ธฐ๋ก์„ ์„ธ์šด ๋ชจ๋ธ์„ ํ›ˆ๋ จ์‹œ์ผฐ ๋‹ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๊ฐ€ ํ•™๊ต์—์„œ ์ˆ˜ํ•™์„ ๋ฐฐ์šฐ๋Š” ๋ฐฉ์‹๊ณผ ๋น„์Šทํ•˜์ฃ ? ์ตœ์ข… ์ •๋‹ต์„ ๋งžํžŒ AI์˜ ๋“ฑ์„ ๋‘๋“œ๋ฆฌ๋Š” ๋Œ€์‹ , ์ด์ œ ์˜ฌ๋ฐ”๋ฅธ ๋‹จ๊ณ„๋ฅผ ๋ฐŸ์„ ๋•Œ๋งˆ๋‹ค ๋ณด์ƒ์„ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค. ๋งˆ์น˜ ์ˆ˜ํ•™ ์„ ์ƒ๋‹˜์ด ์ตœ์ข… ์ •๋‹ต๋ฟ๋งŒ ์•„๋‹ˆ๋ผ ํ•™์Šต ๊ณผ์ •๊นŒ์ง€ ์‹ ๊ฒฝ ์จ์ฃผ๋Š” ๊ฒƒ์ฒ˜๋Ÿผ์š”. ์ด๋ฅผ '๊ณผ์ • ๊ฐ๋…(Process Supervision)'์ด๋ผ๊ณ  ๋ถ€๋ฅด๋Š”๋ฐ, ์ด๋Š” ํš๊ธฐ์ ์ธ ๋ณ€ํ™”์ž…๋‹ˆ๋‹ค. ๐Ÿงฎ๐Ÿ’ก
* ๋…ผ๋ฌธ์˜ ์ œ๋ชฉ์€ "Letโ€™s Verify Step by Step"์ž…๋‹ˆ๋‹ค. (์ž‘๋…„์— CoT - '์ƒ๊ฐ์˜ ์‚ฌ์Šฌ' ๋ถ์„ ์ผ์œผ์ผฐ๋˜ ๋งˆ๋ฒ•์˜ ํ‘œํ˜„ Step by Step ์ž…๋‹ˆ๋‹ค)

https://cdn.openai.com/improving-mathematical-reasoning-with-process-supervision/Lets_Verify_Step_by_Step.pdf?fbclid=IwAR1LojRtgWmMYERWatOlHsD_9AAxalBPDd2i9WiDfRwV_F_Scm3eT2IUB1k
์ด๊ฒƒ์€ ๋‹จ์ˆœํžˆ ์ˆ˜ํ•™ ๋ฌธ์ œ๋ฅผ ๋” ์ž˜ ํ‘ธ๋Š” ๊ฒƒ๋งŒ์ด ์•„๋‹™๋‹ˆ๋‹ค. ํ”„๋กœ์„ธ์Šค ๊ฐ๋…์€ ๋ชจ๋ธ์ด ์‚ฌ๋žŒ์ด ์Šน์ธํ•œ ํ”„๋กœ์„ธ์Šค๋ฅผ ๋”ฐ๋ผ ์šฐ๋ฆฌ์™€ ๋” ๋น„์Šทํ•˜๊ฒŒ ์ƒ๊ฐํ•˜๋„๋ก ๋„์™€์ค๋‹ˆ๋‹ค. ๋งˆ์น˜ ์ธ๊ณต์ง€๋Šฅ์˜ ๋งˆ์Œ์„ ๋“ค์—ฌ๋‹ค๋ณด๊ณ  ์ž‘๋™ํ•˜๋Š” ๊ณผ์ •์„ ์ง€์ผœ๋ณด๋Š” ๊ฒƒ๊ณผ ๋น„์Šทํ•ฉ๋‹ˆ๋‹ค. ์ด ๋ฐฉ๋ฒ•์€ ๋งค์šฐ ์‹ ๋ขฐํ•  ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์œผ๋กœ ์ž…์ฆ๋˜์—ˆ์œผ๋ฉฐ, AI๋ฅผ ๋” ์ž˜ ์ดํ•ดํ•˜๊ณ  ์šฐ๋ฆฌ์˜ ๋ชฉํ‘œ์— ๋ถ€ํ•ฉํ•˜๋„๋ก ๋งŒ๋“ค ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค.
์ด๊ฒƒ์€ AI์˜ ์†๋„๋ฅผ ๋Šฆ์ถ”๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ ์˜คํžˆ๋ ค ๋” ๋‚˜์€ ์„ฑ๋Šฅ์„ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค! ์•ˆ์ „๊ณผ ์„ฑ๋Šฅ ๋ชจ๋‘์—์„œ ์œˆ์œˆ์ธ ์…ˆ์ด์ฃ .
๋งค์šฐ ๊นŒ๋‹ค๋กœ์šด ์ˆ˜ํ•™ ๋ฌธ์ œ๋ฅผ ํ…Œ์ŠคํŠธํ•œ ๊ฒฐ๊ณผ, '๊ณผ์ • ๊ฐ๋…ํ˜•' AI๊ฐ€ '๊ฒฐ๊ณผ ๊ฐ๋…ํ˜•' AI๋ณด๋‹ค ์ „๋ฐ˜์ ์œผ๋กœ ๋” ๋›ฐ์–ด๋‚œ ์„ฑ๋Šฅ์„ ๋ณด์˜€์Šต๋‹ˆ๋‹ค. ์‹ฌ์ง€์–ด ๋Œ€๋ถ€๋ถ„์˜ ์‹œ๋„๊ฐ€ ์‹คํŒจ๋กœ ๋Œ์•„๊ฐ„ ์–ด๋ ค์šด ์‚ผ๊ฐํ˜• ๋ฌธ์ œ์— ๋Œ€ํ•œ ์œ ํšจํ•œ ํ•ด๊ฒฐ์ฑ…์„ ์ฐพ์•„๋‚ด๊ธฐ๋„ ํ–ˆ์Šต๋‹ˆ๋‹ค! ๐Ÿ˜ฎ
ํ•˜์ง€๋งŒ ์ด ์„ฑ๊ณผ๋ฅผ ์นญ์ฐฌํ•˜๋Š”๋ฐ๋งŒ ๊ทธ์น˜๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ ์ง„์ง€ํ•˜๊ฒŒ ๊ณ ๋ฏผํ•ด์•ผ ํ•  ๋ถ€๋ถ„์ด ์žˆ์Šต๋‹ˆ๋‹ค. ๐Ÿค”
์ด ๋ฐฉ๋ฒ•์„ ๋‹ค๋ฅธ ๋ฌธ์ œ ํ•ด๊ฒฐ ์ž‘์—…์—์„œ AI๋ฅผ ํ›ˆ๋ จํ•˜๋Š” ๋ฐ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ์„๊นŒ์š”? ๊ทธ๋ฆฌ๊ณ  ์ด๊ฒƒ์ด AI์˜ ์œค๋ฆฌ์™€ ํˆฌ๋ช…์„ฑ์„ ์–ด๋–ป๊ฒŒ ๋ณ€ํ™”์‹œํ‚ฌ ์ˆ˜ ์žˆ์„๊นŒ์š”? AI์˜ ์˜์‚ฌ๊ฒฐ์ •์„ ๋” ์ž˜ ์ดํ•ดํ•˜๊ณ  ๊ทœ์ œํ•˜๋Š” ๋ฐ ๋„์›€์ด ๋  ์ˆ˜ ์žˆ์„๊นŒ์š”?

AI๊ฐ€ ์ด ์ •๋„๋กœ ๋ฐœ์ „ํ•˜๋ฉด ์ผ๋ถ€ ์ง์—…์ด ์˜ํ–ฅ์„ ๋ฐ›์„ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์–ด๋–ป๊ฒŒ ๋Œ€์‘ํ•ด์•ผ ํ• ๊นŒ์š”? ํŽธ๊ฒฌ์€ ์–ด๋–ป๊ฒŒ ํ• ๊นŒ์š”? ์ด ๋ฐฉ๋ฒ•์ด ํŽธ๊ฒฌ์„ ์‹๋ณ„ํ•˜๊ณ  ์ˆ˜์ •ํ•˜๋Š” ๋ฐ ๋„์›€์ด ๋  ์ˆ˜ ์žˆ์„๊นŒ์š”?
๊ทธ๋ฆฌ๊ณ  ์ค‘์š”ํ•œ ์งˆ๋ฌธ์ด ํ•˜๋‚˜ ๋” ์žˆ์Šต๋‹ˆ๋‹ค. AI๊ฐ€ ์šฐ๋ฆฌ๋ณด๋‹ค ์ถ”๋ก ์— ๋” ๋Šฅ์ˆ™ํ•ด์ง„๋‹ค๋ฉด, ์šฐ๋ฆฌ๊ฐ€ ํ•จ๊ป˜ ์ผํ•˜๋Š” ๋ฐฉ์‹์ด ์–ด๋–ป๊ฒŒ ๋ฐ”๋€”๊นŒ์š”?
๋งˆ์ง€๋ง‰์œผ๋กœ, ์ด ๋ฐฉ๋ฒ•์€ ์ธ๊ฐ„์ด ์Šน์ธํ•œ ํ”„๋กœ์„ธ์Šค๋ฅผ ๋”ฐ๋ฅด๊ธฐ ๋•Œ๋ฌธ์— AI์˜ ์ฐฝ์˜์„ฑ์ด๋‚˜ ํ˜์‹  ์ž ์žฌ๋ ฅ์„ ์ œํ•œํ•  ์ˆ˜ ์žˆ์„๊นŒ์š”?
๊ณ ๋ คํ•ด์•ผ ํ•  ์‚ฌํ•ญ์ด ๋งŽ์ง€๋งŒ ํ•œ ๊ฐ€์ง€ ํ™•์‹คํ•œ ๊ฒƒ์€ ์žˆ์Šต๋‹ˆ๋‹ค: AI๋Š” ์ ์  ๋” ๋˜‘๋˜‘ํ•ด์ง€๊ณ  ์žˆ์œผ๋ฉฐ, ์ด๋ฅผ ํ™œ์šฉํ•˜๋Š” ๊ฒƒ์€ ์šฐ๋ฆฌ์˜ ๋ชซ์ด๋ผ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. #AI #OpenAI #ํ”„๋กœ์„ธ์Šค๊ฐ๋… #MathWhizAI ๐Ÿค–๐Ÿ‘ฉโ€๐Ÿซ๐Ÿ”ข
----
์ €์ž์— ์–ผ๋ผ์ธ๋จผํŠธํŒ€์˜ Jan Leike๊ฐ€ ์žˆ๋Š” ๊ฒƒ์ด ํฅ๋ฏธ๋กญ๊ตฐ์š”.
Jan Leike: https://twitter.com/janleike/status/1663977494058520576?fbclid=IwAR2uKE0yDJOc9V_2Ihh0AJY56TrFPs-BcAhG4kPBHmwR7fFaWrlmAtVq_nI
Leike๋Š” ์ด๊ฒƒ์ด ๊ธฐ๋ณธ์ ์œผ๋กœ AI๊ฐ€ ์šฐ๋ฆฌ๊ฐ€ ์›ํ•˜๋Š” ๋ฐฉ์‹์œผ๋กœ ์ž‘๋™ํ•˜๋„๋ก ํ•˜๋Š” '์ •๋ ฌ'์ด๋ผ๋Š” ์ธก๋ฉด์—์„œ ํฐ ๋‰ด์Šค๊ฐ€ ๋  ๊ฒƒ์ด๋ผ๊ณ  ์ƒ๊ฐํ•˜๋Š”๊ตฐ์š”.
AI ์—…๊ณ„์—์„œ๋Š” '์ •๋ ฌ ์„ธ๊ธˆ'์ด๋ผ๊ณ  ๋ถˆ๋ฆฌ๋Š” ๊ฒƒ, ์ฆ‰ AI๋ฅผ ์ •๋ ฌํ•ด์„œ ์ž‘๋™ํ•˜๊ฒŒ ๋งŒ๋“ค๋ฉด AI์˜ ์„ฑ๋Šฅ์ด ๋–จ์–ด์งˆ ์ˆ˜ ์žˆ๋‹ค๋Š” ์šฐ๋ ค๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋Ÿฐ๋ฐ ์ด ์‚ฌ๋ก€์˜ ๊ฒฝ์šฐ, AI๋Š” ์‹ค์ œ๋กœ ๋ฉด๋ฐ€ํ•œ ๊ฐ๋…์„ ๋ฐ›์•˜์„ ๋•Œ ๋” ์ž˜ ์ˆ˜ํ–‰ํ–ˆ๋‹ค๋Š” ๊ฒƒ์ด์ฃ .
Leike ์ž์‹ ๋„ ์ด ์‚ฌ์‹ค์— ๋†€๋ž๋‚˜ ๋ด…๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋” ํฐ AI ๋ชจ๋ธ์ด๋‚˜ ๋‹ค๋ฅธ ์˜์—ญ์—์„œ๋„ ๊ฐ™์€ ์ผ์ด ์ผ์–ด๋‚ ์ง€ ํ™•์‹ ํ•˜์ง€ ๋ชปํ–ˆ์ง€๋งŒ, ์ง€๊ธˆ์€ ํ™•์‹คํžˆ ์ด ์•„์ด๋””์–ด์— ๋Œ€ํ•ด ๋” ๊ฐœ๋ฐฉ์ ์ด๊ตฐ์š”.
๊ทธ๋Š” ๋˜ํ•œ ์ด๊ฒƒ์ด ์ •๋ ฌ ์—ฐ๊ตฌ์— ํฐ ์˜๋ฏธ๋ฅผ ๊ฐ€์งˆ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. AI์˜ ์ž‘๋™ ๋ฐฉ์‹์„ ๋” ์ž˜ ์ดํ•ดํ•  ์ˆ˜ ์žˆ๋‹ค๋ฉด AI๊ฐ€ ๋” ์ž˜ ์ž‘๋™ํ•˜๋„๋ก ํ•  ์ˆ˜ ์žˆ์ง€ ์•Š์„๊นŒ์š”?
* PRM800K: A Process Supervision Dataset: https://github.com/openai/prm800k
----
* ํฌ์ŠคํŒ… ์ž‘์„ฑ ๊ณผ์ •: https://chat.openai.com/share/28921572-7b11-496c-a533-9898733d6577?fbclid=IwAR0qLLSN6H3aaiamCK-XFeZfCIa_wEXwBUBtlMjn9dN7h7F7UQH9f6l4qEQ
๊ณผ์—ฐ ์ด๋ ‡๊ฒŒ ๊ธ€์“ฐ๊ธฐ๋ฅผ ํ•˜๋Š” ๊ฒƒ์€ ์ œ๊ฐ€ ์ผ๋‹ค๊ณ  ๋ณผ ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์ธ์ง€, ๊ณผ์—ฐ ์ €๋Š” ์›๋ฌธ์˜ ๋‚ด์šฉ์„ ์ฝ์—ˆ๋‹ค๊ณ  ๋ณผ ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์ธ์ง€... ๊ธ€์Ž„์š”.
๋งˆ์น˜ ์›๋ฌธ์„ ์ฝ์€ ๋‹ค๋ฅธ ์‚ฌ๋žŒ์ด ์“ด ํฌ์ŠคํŒ…์„ ๋ณด๊ณ , ๊ทธ๊ฑธ ํ†ตํ•ด ๋Œ€๊ฐ• ํŒŒ์•…ํ•˜๋Š” ๋А๋‚Œ์ด๋ž„๊นŒ์š”. ์ •๋ณด์˜ ํก์ˆ˜๋ผ๋Š”๊ฒŒ ๋Š˜ ์ด๋Ÿฐ ๋‹จ๊ณ„๊ฐ€ ์žˆ๋Š” ๊ฒƒ์ด๊ธด ํ•˜์ง€๋งŒ... ๋ญ”๊ฐ€์˜ ๊ฐˆ๋ฆผ๊ธธ์ด๊ตฌ๋‚˜ ์‹ถ์Šต๋‹ˆ๋‹ค. ์—ฌ๊ธฐ์„œ ๋” ๋“ค์–ด์„œ ์ œ๋Œ€๋กœ ์ฝ๊ณ  ์ƒ๊ฐํ• ์ง€ ์•„๋‹ˆ๋ฉด ๊ฒŒ์œผ๋ฆ„์„ ํ”ผ์šฐ๊ณ  ๋ง์ง€. ๊ทธ๋Ÿฐ๋ฐ ์ œ๋Œ€๋กœ ์ฝ๊ณ  ์ƒ๊ฐํ•œ๋‹ค์† ์น˜๋”๋ผ๋„ ์‹œ๊ฐ„์ด ์ง€๋‚˜ ๋‚จ๋Š” ๊ฒƒ์ด ์ด๋Ÿฐ ๊ฒŒ์œผ๋ฅธ ํฌ์ŠคํŒ… ๋Œ€๋น„ ๊ณผ์—ฐ ๋ญ๊ฐ€ ๋” ์žˆ์„ ๊ฒƒ์ธ๊ฐ€... ์ด๊ฒŒ ๊ณ ๋ฏผ ์ง€์ ์ž…๋‹ˆ๋‹ค.
* ๋ญ”๊ฐ€ ์ฐœ์ฐœํ•ด์„œ ๊ฒฐ๊ตญ ๋ธ”๋กœ๊ทธ ์›๋ฌธ์€ ๋‹ค ์ฝ์—ˆ์Šต๋‹ˆ๋‹ค. ์ฝ๊ณ  ๋‚˜์„œ ์ด ํฌ์ŠคํŒ…์˜ ๋ณธ๋ฌธ์— ์ด๋ ‡๋‹คํ•  ์ˆ˜์ •์„ ํ•˜์ง€ ์•Š๋Š” ๊ฒƒ๋„ ์•„์ด๋Ÿฌ๋‹ˆํ•ฉ๋‹ˆ๋‹ค. ๋…ผ๋ฌธ๋„ ๋‘˜๋Ÿฌ๋ด์•ผ์ฃ .
What if we apply this methodology into other areas? Maybe at some point, human need to learn from AI about how to learn...
Japan Goes All In: Copyright Doesnโ€™t Apply To AI Training
์ผ๋ณธ ์ •๋ถ€๋Š” ์ธ๊ณต์ง€๋Šฅ ํ•™์Šต์— ์‚ฌ์šฉ๋˜๋Š” ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•ด์„œ ์ €์ž‘๊ถŒ์„ ๊ฐ•์ œํ•˜์ง€ ์•Š๊ฒ ๋‹ต๋‹ˆ๋‹ค.
"์ผ๋ณธ ์ •๋ถ€๋Š” ์ตœ๊ทผ ๋†€๋ž๊ฒŒ๋„ AI ํ•™์Šต์— ์‚ฌ์šฉ๋˜๋Š” ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•ด ์ €์ž‘๊ถŒ์„ ๊ฐ•์ œํ•˜์ง€ ์•Š๊ฒ ๋‹ค๋Š” ์ž…์žฅ์„ ์žฌํ™•์ธํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด ์ •์ฑ…์€ "๋น„์˜๋ฆฌ ๋ชฉ์ ์ด๋“  ์ƒ์—…์  ๋ชฉ์ ์ด๋“ , ๋ณต์ œ๊ฐ€ ์•„๋‹Œ ๋‹ค๋ฅธ ํ–‰์œ„์ด๋“ , ๋ถˆ๋ฒ• ์‚ฌ์ดํŠธ ๋“ฑ์—์„œ ์–ป์€ ์ฝ˜ํ…์ธ ์ด๋“  ์ƒ๊ด€์—†์ด" AI๊ฐ€ ๋ชจ๋“  ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ๋„๋ก ํ—ˆ์šฉํ•ฉ๋‹ˆ๋‹ค. ์ผ๋ณธ ๋ฌธ๋ถ€๊ณผํ•™์„ฑ ์žฅ๊ด€์ธ ๋‚˜๊ฐ€์˜ค์นด ์ผ€์ด์ฝ”๋Š” ํ˜„์ง€ ํšŒ์˜์—์„œ ์ผ๋ณธ ๋ฒ•๋ฅ ์ด AI ๋ฐ์ดํ„ฐ ์„ธํŠธ์— ์‚ฌ์šฉ๋œ ์ €์ž‘๊ถŒ์ด ์žˆ๋Š” ์ž๋ฃŒ๋ฅผ ๋ณดํ˜ธํ•˜์ง€ ์•Š์„ ๊ฒƒ์ด๋ผ๊ณ  ๋งํ•˜๋ฉฐ ๋Œ€๋‹ดํ•œ ์ž…์žฅ์„ ํ™•์ธํ–ˆ์Šต๋‹ˆ๋‹ค."
๊ทธ๋Ÿฐ๋ฐ ์ด์— ๋Œ€์‘ํ•˜๋Š” ๊ตญ๋‚ด ๊ธฐ์‚ฌ๋Š” ๋ชป์ฐพ๊ฒ ๋„ค์š”. (์žˆ๋Š”๋ฐ ์ œ๊ฐ€ ์ฐพ์ง€ ๋ชปํ•˜๋Š” ๊ฑธ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค.)

https://technomancers.ai/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/?fbclid=IwAR0jWJD6EvO9eE6hTdjqUBOuGkTXVYIVZ1iuP_cLgqj7_Sz2xs8K4tKgyTM

Huge decision for Japan. Smart goverment.
Continuous Learning_Startup & Investment
honeycomb.io/blog/hard-stuff-nobody-talks-about-llm?fbclid=IwAR1pDrPdN7Vu9PSEwK6aD_lWCZMoZpbf5cQjI0nZ1MDWQfJUNO8ZUM27G74_aem_th_AUxsqnFXcjp2E4Hl4UUDPEp6EK_W_tS9Vkgp9aXhW0Kq5cKDefW90jh2ARchnFjpies&mibextid=Zxz2cZ
Builderโ€™s point of view about LLM.

The article discusses the challenges of building products with large language models (LLMs). These models are trained on massive datasets of text and code, and can be used for a variety of tasks, such as generating text, translating languages, and writing different kinds of creative content. However, building products with LLMs is not without its challenges. One challenge is that LLMs are often very large and complex, which can make them difficult to train and deploy. Another challenge is that LLMs can be difficult to control, and can sometimes generate output that is offensive or harmful. Finally, LLMs can be expensive to use, and there is a risk that they could be used for malicious purposes.

Despite these challenges, the author believes that LLMs have the potential to revolutionize the way we build products. He argues that LLMs can be used to create more personalized and engaging experiences for users, and that they can be used to solve complex problems that were previously intractable. He concludes by calling for more research and development into the use of LLMs in products.

Here are some of the specific challenges mentioned in the article:

Training and deploying LLMs is expensive and time-consuming. LLMs require massive datasets of text and code to train, and can take weeks or even months to train. Once they are trained, they can be difficult to deploy, as they require specialized hardware and software.
LLMs can be difficult to control. LLMs are trained on massive datasets of text, and can sometimes generate output that is offensive or harmful. This can be a challenge for developers, who need to find ways to control the output of LLMs.
LLMs can be expensive to use. LLMs require specialized hardware and software, which can be expensive to purchase and maintain. Additionally, LLMs can be expensive to train, as they require massive datasets of text and code.
There is a risk that LLMs could be used for malicious purposes. LLMs are powerful tools, and could be used for malicious purposes, such as generating fake news or creating deepfakes. It is important to be aware of these risks when developing products with LLMs.
10 common mistakes startups make that result in slower growth:

1. Focusing on middle of the funnel when the bottleneck is top of the funnel. Look at reps calendars. Are they busy with multiple customer calls every day? If not, you have a demand (and/or p/m fit) bottleneck.

2. Mass email sequencing people who have never heard of you or your company. You need to be more creative with outbound.

3. Incentivizing the wrong behavior. Most commonly, top of funnel metrics like MQLโ€™s and SQLโ€™s that produce low conversion rates and low ACVโ€™s. โ€œShow me the incentive and Iโ€™ll show you the outcome.โ€ - Charlie Munger

4. Quarterly quotas. Almost every startup should be on monthly quotas. The exception being high 6 figure ACVs with >90 day sales cycles. The consequence of quarterly quotas is lost urgency 2 out of 3 months and less revenue.

5. Underinvesting or investing too late in revenue operations.

6. Underinvesting or investing too late in implementation. The first 30-60 days are the most critical in determining the long term success of a customer. Do whatever it takes to deliver a delightful first impression.

7. Underinvesting or investing too late in revenue expansion in the customer base.

8. Never meeting customers in person. Founders and revenue leaders should be doing this regularly (yes, even if your customers are small).

9. Over-indexing on domain experience when hiring a VP of Sales. Or, less common, but an even worse mistake, hiring from a big company.

10. Allowing poor performers, especially those in leadership positions, to remain at the business for too long.

https://www.linkedin.com/posts/sam-blond-791026b_10-common-mistakes-startups-make-that-result-activity-7070149376043573250-QLTB?utm_source=share&utm_medium=member_desktop
Continuous Learning_Startup & Investment
10 common mistakes startups make that result in slower growth: 1. Focusing on middle of the funnel when the bottleneck is top of the funnel. Look at reps calendars. Are they busy with multiple customer calls every day? If not, you have a demand (and/or p/mโ€ฆ
์Šคํƒ€ํŠธ์—…์˜ ์„ฑ์žฅ์ด ๋‘”ํ™”๋˜๋Š” ๋ฐ ์ฃผ์š”ํ•œ ์—ญํ• ์„ ํ•˜๋Š” 10๊ฐ€์ง€ ์ผ๋ฐ˜์ ์ธ ์‹ค์ˆ˜
Sam Blond Partner at Founders Fund, Ex CRO in Brex

1. ๊น”๋•Œ๊ธฐ ์ƒ๋‹จ์— ๋ณ‘๋ชฉ ์ง€์ ์ด ์žˆ์Œ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ๊น”๋•Œ๊ธฐ ์ค‘๊ฐ„์—๋งŒ ์ง‘์ค‘ํ•˜๋Š” ๊ฒฝ์šฐ. ๋‹ด๋‹น์ž์˜ ์ผ์ •์„ ์ฒดํฌํ•ด๋ณด์„ธ์š”. ๋งค์ผ ์ˆ˜๋งŽ์€ ๊ณ ๊ฐ ์ „ํ™”๋กœ ๋ฐ”์˜๊ฒŒ ์ง€๋‚ด๋‚˜์š”? ๊ทธ๋ ‡์ง€ ์•Š๋‹ค๋ฉด, ์ˆ˜์š”(๋˜๋Š” ์ œํ’ˆ/์‹œ์žฅ ์ ํ•ฉ์„ฑ)์— ๋ณ‘๋ชฉ์ด ์ƒ๊ธฐ๊ณ  ์žˆ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

2. ๋ณธ์ธ์˜ ํšŒ์‚ฌ๋‚˜ ๋ณธ์ธ์˜ ํšŒ์‚ฌ์— ๋Œ€ํ•ด ์•Œ์ง€ ๋ชปํ•˜๋Š” ์‚ฌ๋žŒ๋“ค์—๊ฒŒ ๋Œ€๋Ÿ‰์œผ๋กœ ์ด๋ฉ”์ผ์„ ๋ณด๋‚ด๋Š” ๊ฒฝ์šฐ. ์•„์›ƒ๋ฐ”์šด๋“œ ์ „๋žต์— ๋” ๋งŽ์€ ์ฐฝ์˜์„ฑ์„ ๋ฐœํœ˜ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

3. ๋ถ€์ ์ ˆํ•œ ํ–‰๋™์— ์ธ์„ผํ‹ฐ๋ธŒ๋ฅผ ์ œ๊ณตํ•˜๋Š” ๊ฒฝ์šฐ. ๊ฐ€์žฅ ํ”ํ•œ ์˜ˆ๋Š”, ๋‚ฎ์€ ์ „ํ™˜์œจ๊ณผ ๋‚ฎ์€ ACV๋ฅผ ์ƒ์„ฑํ•˜๋Š” MQL ๋ฐ SQL ๊ฐ™์€ ํผ๋„ ์ƒ๋‹จ ์ง€ํ‘œ์ž…๋‹ˆ๋‹ค. "์ธ์„ผํ‹ฐ๋ธŒ๋ฅผ ์„ค์ •ํ•˜๋ฉด ๊ฒฐ๊ณผ๊ฐ€ ๋”ฐ๋ผ์˜ต๋‹ˆ๋‹ค." - ์ฐฐ๋ฆฌ ๋ฉ๊ฑฐ

4. ๋ถ„๊ธฐ๋ณ„ ๋ชฉํ‘œ ์„ค์ •. ๋Œ€๋ถ€๋ถ„์˜ ์Šคํƒ€ํŠธ์—…์€ ์›”๋ณ„ ๋ชฉํ‘œ๋ฅผ ์„ค์ •ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๋‹จ, ํŒ๋งค ์ฃผ๊ธฐ๊ฐ€ 90์ผ์„ ์ดˆ๊ณผํ•˜๋Š” ๊ณ ์•ก์˜ ACV๋Š” ์˜ˆ์™ธ์ž…๋‹ˆ๋‹ค. ๋ถ„๊ธฐ๋ณ„ ๋ชฉํ‘œ์˜ ๊ฒฐ๊ณผ๋Š” 3๊ฐœ์›” ์ค‘ 2๊ฐœ์›” ๋™์•ˆ ๊ธด๊ธ‰์„ฑ์„ ์žƒ๊ณ  ์ˆ˜์ต์ด ๊ฐ์†Œํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.

5. ์ˆ˜์ต ์šด์˜์— ๋Œ€ํ•œ ํˆฌ์ž๊ฐ€ ๋ถ€์กฑํ•˜๊ฑฐ๋‚˜ ๋„ˆ๋ฌด ๋Šฆ๊ฒŒ ํˆฌ์žํ•˜๋Š” ๊ฒฝ์šฐ.

6. ๊ณผ์†Œ ํˆฌ์žํ•˜๊ฑฐ๋‚˜ ์‹คํ–‰์— ๋„ˆ๋ฌด ๋Šฆ๊ฒŒ ์ฐฉ์ˆ˜ํ•˜๋Š” ๊ฒฝ์šฐ. ๊ณ ๊ฐ์˜ ์žฅ๊ธฐ์ ์ธ ์„ฑ๊ณต์„ ๊ฒฐ์ •ํ•˜๋Š” ๊ฐ€์žฅ ์ค‘์š”ํ•œ ์‹œ๊ธฐ๋Š” ์ฒ˜์Œ 30~60์ผ์ž…๋‹ˆ๋‹ค. ์ข‹์€ ์ฒซ์ธ์ƒ์„ ๋‚จ๊ธฐ๊ธฐ ์œ„ํ•ด ์ตœ์„ ์„ ๋‹คํ•˜์‹ญ์‹œ์˜ค.

7. ๊ณ ๊ฐ ๊ธฐ๋ฐ˜ ๋งค์ถœ ํ™•๋Œ€์— ๋Œ€ํ•œ ํˆฌ์ž๊ฐ€ ๋ถ€์กฑํ•˜๊ฑฐ๋‚˜ ๋„ˆ๋ฌด ๋Šฆ๊ฒŒ ํ•˜๋Š” ๊ฒฝ์šฐ.

8. ๊ณ ๊ฐ์„ ์ง์ ‘ ๋งŒ๋‚˜์ง€ ์•Š๋Š” ๊ฒฝ์šฐ. ์ฐฝ์—…์ž์™€ ๋งค์ถœ ๋ฆฌ๋”๋Š” ์ •๊ธฐ์ ์œผ๋กœ ๊ณ ๊ฐ์„ ๋งŒ๋‚˜์•ผ ํ•ฉ๋‹ˆ๋‹ค(๊ณ ๊ฐ์ด ์†Œ๊ทœ๋ชจ์ผ์ง€๋ผ๋„).

9. ์˜์—… ๋ถ€์‚ฌ์žฅ์„ ์ฑ„์šฉํ•  ๋•Œ ๋„๋ฉ”์ธ ๊ฒฝํ—˜์„ ๊ณผ๋„ํ•˜๊ฒŒ ์ค‘์š”์‹œํ•˜๋Š” ๊ฒฝ์šฐ. ๋˜๋Š” ํ”ํ•˜์ง€ ์•Š์ง€๋งŒ ๋”์šฑ ์‹ฌ๊ฐํ•œ ์‹ค์ˆ˜์ธ ๋Œ€๊ธฐ์—… ์ถœ์‹ ์„ ์ฑ„์šฉํ•˜๋Š” ๊ฒฝ์šฐ

10. ์„ฑ๊ณผ๊ฐ€ ์ €์กฐํ•œ ์ง์›, ํŠนํžˆ ๋ฆฌ๋”์‹ญ ์ง์ฑ…์— ์žˆ๋Š” ์ง์›์„ ๋„ˆ๋ฌด ์˜ค๋ž˜ ํšŒ์‚ฌ์— ๋จธ๋ฌผ๊ฒŒ ํ•˜๋Š” ๊ฒฝ์šฐ
โค1
Continuous Learning_Startup & Investment
Google is investing in Runway, a New Yorkโ€“based startup that lets customers generate video from text descriptions using artificial intelligence it pioneered, at a valuation of around $1.5 billion including the new capital, according to two people familiarโ€ฆ
I missed a great point here.
As Microsoft bankrolls AI startup OpenAIโ€”including footing its enormous cloud bill as part of a complex arrangementโ€”Microsoftโ€™s rivals have scrambled to win over other AI customers. Google has begun to invest in AI startups that are or could become its cloud customers, following a similar playbook it used to obtain customers in other industries. Earlier this year, for instance, Google invested $400 million in Anthropic, one of the primary startup challengers to OpenAI, The Information previously reported. Google Cloud said in February it had become Anthropicโ€™s โ€œpreferredโ€ provider.

News of the Runway deal comes after Google Cloud announced it had signed up two other key AI startups: text-to-image provider Midjourney and chatbot app Character AI. Character, founded by former Google researchers who developed some of the companyโ€™s key AI technology, previously was a key cloud customer of Oracle, The Information reported. It isnโ€™t clear whether Google took stakes in the companies and itโ€™s too soon to tell whether these deals will help Google catch up to bigger cloud rivals AWS and Microsoft. See our Cloud Database.

Notably, OpenAI has employed its own startup-investment strategy to win over future customers even as it takes capital and cloud credits from Microsoft. The startup, which raised a $175 million fund to back startups, dangles special incentives to founders whose startups pay for access to its large-language models, including discounts on licensing and early access to new technologies, The Information previously reported.

Luring AI startups with cloud credits could help Google beat out venture capital firms vying for stakes in the companies. Oracle also offers hundreds of thousands of dollarsโ€™ worth of computing credits as an incentive for AI startups to rent the companyโ€™s cloud servers, The Information earlier reported.
Microsoft๊ฐ€ ๋ณต์žกํ•œ ๊ณ„์•ฝ์˜ ์ผํ™˜์œผ๋กœ ๋ง‰๋Œ€ํ•œ ํด๋ผ์šฐ๋“œ ๋น„์šฉ์„ ๋ถ€๋‹ดํ•˜๋Š” ๋“ฑ AI ์Šคํƒ€ํŠธ์—… OpenAI์— ์ž๊ธˆ์„ ์ง€์›ํ•˜๋ฉด์„œ, ๊ฒฝ์Ÿ์‚ฌ๋“ค์€ ๋‹ค๋ฅธ AI ๊ณ ๊ฐ์„ ํ™•๋ณดํ•˜๊ธฐ ์œ„ํ•ด ๋ถ„์ฃผํ•˜๊ฒŒ ์›€์ง์ด๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€์€ ๋‹ค๋ฅธ ์—…๊ณ„์—์„œ ๊ณ ๊ฐ์„ ํ™•๋ณดํ•  ๋•Œ ์‚ฌ์šฉํ•œ ์œ ์‚ฌํ•œ ์ „๋žต์— ๋”ฐ๋ผ ํด๋ผ์šฐ๋“œ ๊ณ ๊ฐ์ด๊ฑฐ๋‚˜ ํด๋ผ์šฐ๋“œ ๊ณ ๊ฐ์ด ๋  ์ˆ˜ ์žˆ๋Š” AI ์Šคํƒ€ํŠธ์—…์— ํˆฌ์žํ•˜๊ธฐ ์‹œ์ž‘ํ–ˆ์Šต๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด, ์˜ฌํ•ด ์ดˆ ๊ตฌ๊ธ€์€ OpenAI์— ๋„์ „ํ•˜๋Š” ์ฃผ์š” ์Šคํƒ€ํŠธ์—… ์ค‘ ํ•˜๋‚˜์ธ ์•คํŠธ๋กœํ”ฝ์— 4์–ต ๋‹ฌ๋Ÿฌ๋ฅผ ํˆฌ์žํ–ˆ๋‹ค๊ณ  ๋”์ธํฌ๋ฉ”์ด์…˜์€ ์ด์ „์— ๋ณด๋„ํ•œ ๋ฐ” ์žˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€ ํด๋ผ์šฐ๋“œ๋Š” 2์›”์— ์•คํŠธ๋กœํ”ฝ์˜ "์šฐ์„ " ์ œ๊ณต์—…์ฒด๊ฐ€ ๋˜์—ˆ๋‹ค๊ณ  ๋ฐํ˜”์Šต๋‹ˆ๋‹ค.

๋Ÿฐ์›จ์ด ๊ณ„์•ฝ ์†Œ์‹์€ ๊ตฌ๊ธ€ ํด๋ผ์šฐ๋“œ๊ฐ€ ํ…์ŠคํŠธ-์ด๋ฏธ์ง€ ๋ณ€ํ™˜ ์„œ๋น„์Šค ์ œ๊ณต์—…์ฒด์ธ ๋ฏธ๋“œ์ €๋‹ˆ์™€ ์ฑ—๋ด‡ ์•ฑ์ธ ์บ๋ฆญํ„ฐ AI๋ผ๋Š” ๋‘ ๊ฐœ์˜ ์ฃผ์š” AI ์Šคํƒ€ํŠธ์—…์„ ์˜์ž…ํ–ˆ๋‹ค๊ณ  ๋ฐœํ‘œํ•œ ์ดํ›„์— ๋‚˜์˜จ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ตฌ๊ธ€์˜ ์ฃผ์š” AI ๊ธฐ์ˆ ์„ ๊ฐœ๋ฐœํ•œ ์ „์ง ๊ตฌ๊ธ€ ์—ฐ๊ตฌ์›๋“ค์ด ์„ค๋ฆฝํ•œ Character๋Š” ์ด์ „์— ์˜ค๋ผํด์˜ ์ฃผ์š” ํด๋ผ์šฐ๋“œ ๊ณ ๊ฐ์ด์—ˆ๋‹ค๊ณ  The Information์€ ๋ณด๋„ํ–ˆ์Šต๋‹ˆ๋‹ค. ๊ตฌ๊ธ€์ด ๋‘ ํšŒ์‚ฌ์˜ ์ง€๋ถ„์„ ์ธ์ˆ˜ํ–ˆ๋Š”์ง€ ์—ฌ๋ถ€๋Š” ๋ช…ํ™•ํ•˜์ง€ ์•Š์œผ๋ฉฐ, ์ด๋ฒˆ ๊ฑฐ๋ž˜๊ฐ€ ๊ตฌ๊ธ€์ด ๋” ํฐ ํด๋ผ์šฐ๋“œ ๋ผ์ด๋ฒŒ์ธ AWS์™€ Microsoft๋ฅผ ๋”ฐ๋ผ์žก๋Š” ๋ฐ ๋„์›€์ด ๋ ์ง€ ๋งํ•˜๊ธฐ๋Š” ์•„์ง ์ด๋ฅด๋‹ค. ํด๋ผ์šฐ๋“œ ๋ฐ์ดํ„ฐ๋ฒ ์ด์Šค๋ฅผ ์ฐธ์กฐํ•˜์„ธ์š”.

ํŠนํžˆ OpenAI๋Š” Microsoft๋กœ๋ถ€ํ„ฐ ์ž๋ณธ๊ณผ ํด๋ผ์šฐ๋“œ ํฌ๋ ˆ๋”ง์„ ๋ฐ›์œผ๋ฉด์„œ๋„ ๋ฏธ๋ž˜ ๊ณ ๊ฐ์„ ํ™•๋ณดํ•˜๊ธฐ ์œ„ํ•ด ์ž์ฒด์ ์ธ ์Šคํƒ€ํŠธ์—… ํˆฌ์ž ์ „๋žต์„ ์ฑ„ํƒํ–ˆ์Šต๋‹ˆ๋‹ค. ์Šคํƒ€ํŠธ์—… ์ง€์›์„ ์œ„ํ•ด 1์–ต 7,500๋งŒ ๋‹ฌ๋Ÿฌ์˜ ํŽ€๋“œ๋ฅผ ์กฐ์„ฑํ•œ ์ด ์Šคํƒ€ํŠธ์—…์€ ๋ผ์ด์„ ์Šค ํ• ์ธ, ์‹ ๊ธฐ์ˆ  ์กฐ๊ธฐ ์•ก์„ธ์Šค ๋“ฑ ์ž์‚ฌ์˜ ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ์— ๋Œ€ํ•œ ์•ก์„ธ์Šค ๋น„์šฉ์„ ์ง€๋ถˆํ•˜๋Š” ์ฐฝ์—…์ž์—๊ฒŒ ํŠน๋ณ„ํ•œ ์ธ์„ผํ‹ฐ๋ธŒ๋ฅผ ์ œ๊ณตํ•œ๋‹ค๊ณ  ๋”์ธํฌ๋ฉ”์ด์…˜์€ ์ด์ „์— ๋ณด๋„ํ•œ ๋ฐ” ์žˆ์Šต๋‹ˆ๋‹ค.

ํด๋ผ์šฐ๋“œ ํฌ๋ ˆ๋”ง์œผ๋กœ AI ์Šคํƒ€ํŠธ์—…์„ ์œ ์ธํ•˜๋ฉด Google์ด ํšŒ์‚ฌ ์ง€๋ถ„์„ ๋†“๊ณ  ๊ฒฝ์Ÿํ•˜๋Š” ๋ฒค์ฒ˜ ์บํ”ผํƒˆ ํšŒ์‚ฌ๋ฅผ ์ด๊ธฐ๋Š” ๋ฐ ๋„์›€์ด ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์˜ค๋ผํด์€ ๋˜ํ•œ AI ์Šคํƒ€ํŠธ์—…์ด ์ž์‚ฌ์˜ ํด๋ผ์šฐ๋“œ ์„œ๋ฒ„๋ฅผ ์ž„๋Œ€ํ•  ๊ฒฝ์šฐ ์ธ์„ผํ‹ฐ๋ธŒ๋กœ ์ˆ˜์‹ญ๋งŒ ๋‹ฌ๋Ÿฌ ์ƒ๋‹น์˜ ์ปดํ“จํŒ… ํฌ๋ ˆ๋”ง์„ ์ œ๊ณตํ•œ๋‹ค๊ณ  The Information์€ ์•ž์„œ ๋ณด๋„ํ•œ ๋ฐ” ์žˆ์Šต๋‹ˆ๋‹ค.