Continuous Learning_Startup & Investment
2.44K 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
Whenever I use an AI product that offers a great user experience, I ask myself: "How the F#๐Ÿคฌ did they do it?"

And if the product is open source, I dive deep into the code, the logic, and of course, the almighty prompt.

Today, I was exploring Bamboo AI, a tool that analyzes data using natural language processing.

Being the curious geek that I am, I dove deep into the prompt to see how it was engineered and to see what types of techniques I could pick up to make my own prompts better.

And in this tweet, I will do my best to reverse engineer the prompt into its building blocks.

Feel free to bookmark this tweet for later reference. I've broken down the prompt into simple pieces for you to replicate if you want.

PS: A (slighlty reduced) snippet of the prompt is attached on the images for reference.
Here's what I found:

1. Role

The prompt starts by declaring a clearly defined role for the AI. Most prompts do this, as this has become a standard best practice.

2. Goal

A clearly defined goal on top of the role allows the AI to act in accordance to it. Nothing extraordinary with this prompt until now.

The problem is that most people stop crafting their prompts here, and then wonder why their outputs are useless and random more often than not.


3. Clearly defined input

The prompt clearly states what is the expected input the AI will receive.

This part of the prompt is often overlooked, but I've found it greatly reduces the randomness of the output.

4. Clearly Defined output

Similarly, clearly going over the expected output in minute detail will help steer the model in the exact direction that you want.

This will allow you to pinpoint exactly what it should do, and on top of that, will reduce the need for revisions.

Again, most people never even get to this point and then wonder why the AI never gets them right.

Of course it doesn't; it won't get you right if you haven't told it what to do.

5. Revisions

The prompt clearly states that revisions are to be expected and the output probably won't be the final one.

Once again, clearly stating what can happen during the actual use of this tool.

6. Input example

On top of clearly defining what input to expect, the prompt also shows an input example.

"Show, don't just tell" is a good principle to keep in mind when prompting.

This will greatly reduce the randomness of the model and make for more accurate outputs.

7. Output example

Stating the examples of the output is equally important.

This will allow the model to pick up on the input -> output pattern and make its answers way more relevant, contextual and useful.

See a pattern here?

Clearly state what the AI should do and what to expect.

Don't leave it to chance if you want your outputs to be reliable and useful.

And now that we have reverse engineered how this prompt works, you will hopefully have ideas on how to improve your own prompts.

I sure did.

https://twitter.com/Luc_AI_Insights/status/1668792631806050304?s=20
https://www.nea.com/blog/4-trends-for-ai-startups-and-generative-ai-companies

Key Insights for Preparing AI-Related Service

#1 Generative AI is changing the rules of company building. Generative AI is a new technology that allows machines to create new content, such as images, videos, and text, that is similar to human-generated content. This technology is changing the way companies are built, and entrepreneurs need to be aware of this trend to stay competitive.

#2 Speed is critical in hiring talent for AI startups. The demand for AI talent is high, and startups need to act fast to hire the right people. Instead of relying on recruiters and culture-fit discussions, startups can take a more straightforward approach to hiring talent.

#3AI startups need to focus on solving real-world problems. AI startups should focus on solving real-world problems, such as improving healthcare, transportation, and education. By focusing on these problems, startups can create value for their customers and make a positive impact on society.

#4 Collaboration is key to success in the AI industry. Collaboration between AI startups, established companies, and academic institutions is essential for success in the AI industry. By working together, companies can share knowledge, resources, and expertise to create innovative solutions.

In summary, entrepreneurs who are preparing AI-related services should be aware of the emerging trends in the AI industry. They should focus on solving real-world problems, act fast in hiring talent, and collaborate with other companies and institutions to create innovative solutions. By following these trends, entrepreneurs can stay competitive and create value for their customers.
https://freebutdeep.substack.com/

์ตœ๊ทผ์— ์•Œ๊ฒŒ ๋œ ๋‰ด์Šค๋ ˆํ„ฐ์ค‘์— ์žฌ์ผ ์žฌ๋ฏธ์žˆ๊ฒŒ ์ฝ๊ณ  ์žˆ๋Š” ๋‰ด์Šค๋ ˆํ„ฐ์ธ๋ฐ์š”. ์‹ค์ œ๋กœ ๋งŒ๋‚˜๋ต™๊ณ  ์ด์•ผ๊ธฐ๋ฅผ ๋‚˜๋ˆ„์–ด๋ด๋„ ํ†ต์ฐฐ์ด ๊ฐ€๋“ํ•˜์‹  ํ›Œ๋ฅญํ•˜์‹  ๋ถ„์ด์‹ญ๋‹ˆ๋‹ค. ์—ฌ๋Ÿฌ ์ฃผ์ œ๋ฅผ ๋‹ค๋ค„์ฃผ์‹œ๋Š”๋ฐ, ๋งŽ์€ ๋ถ„๋“ค๊ป˜ ์•Œ๋ ค์ง€๋ฉด ์ข‹๊ฒ ๋‹ค๋Š” ์ƒ๊ฐ๋“ค์–ด ์กฐ๊ธˆ ๋œฌ๊ธˆ ์—†์ง€๋งŒ ๊ณต์œ ํ•ด๋ด…๋‹ˆ๋‹ค.
์ง„์ž…์žฅ๋ฒฝ์ด ๋ฌด์—‡์ž…๋‹ˆ๊นŒ? ์›์ฒœ๊ธฐ์ˆ ์ด ๋ฌด์—‡์ธ๊ฐ€์š”?

IR ๋˜๋Š” ๋Œ€๊ธฐ์—… ์ž„์› ๋ถ„๋“ค๊ณผ ๋ฏธํŒ… ์‹œ ๊ฐ€๋” ๋“ฃ๋Š” ์งˆ๋ฌธ์ด๋‹ค.

ํŠนํžˆ๋‚˜, ๊ณผ๊ฑฐ ํŠนํ—ˆ๋‚˜ IP ์ค‘์‹ฌ์œผ๋กœ ์ง„์ž…์žฅ๋ฒฝ์„ ์Œ“์•„์˜ค๋˜ ์‹œ์ ˆ์— ๋น„์ฆˆ๋‹ˆ์Šค ํ•˜์‹œ๋Š” ๋ถ„๋“ค์€, ํŠนํžˆ ๋” ๋งŽ์ด ์งˆ๋ฌธํ•˜๋Š” ๋ถ€๋ถ„์ด๊ธฐ๋„ ํ•˜๋‹ค. ๋‚˜๋Š” ๊ณผ๊ฑฐ BCG ์‹œ์ ˆ IP ํ”„๋กœ์ ํŠธ๋„ ํ•ด๋ดค์—ˆ๊ธฐ์—, ๊ทธ ์งˆ๋ฌธ์ด ๊ฐ€์ง€๋Š” ํ•จ์˜๋ฅผ ๋‚˜๋ฆ„ ์ž˜ ์ดํ•ดํ•˜๊ณ  ์žˆ๊ธฐ๋„ ํ•˜๋‹ค.

๋‹ค๋งŒ, ์ง„์ž…์žฅ๋ฒฝ์˜ ์ •์˜๋ฅผ, 1) ํƒ€์‚ฌ ๋Œ€๋น„ ์šฐ๋ฆฌ ์„œ๋น„์Šค๊ฐ€ ์ง€๋‹ˆ๋Š” ์ฐจ๋ณ„์  ๊ฒฝ์Ÿ์šฐ์œ„, 2) ํƒ€์‚ฌ๊ฐ€ ์ง„์ž… ๋ชปํ•˜๊ฒŒ ๋ง‰๋Š”, ์šฐ๋ฆฌ ํšŒ์‚ฌ๋งŒ์˜ ๋…์ ์  ๊ธฐ์ˆ /ํŠนํ—ˆ๋กœ ์ •์˜ํ•˜๋Š” ๊ฒƒ์€, ํ˜„ IT ์—…์—์„œ๋Š” 100% ๋ถ€ํ•ฉํ•˜์ง€๋Š” ์•Š๋Š”๋‹ค.

๋ฏธ๊ตญ์—์„œ ๋งŒ๋‚œ ํ›Œ๋ฅญํ•œ ์„œ๋น„์Šค๋ฅผ ์ผ๊ตฐ ๋ถ„๋“ค์˜ ๊ณตํ†ต์ ์€, ๊ทธ ๋จธ๋ฆฟ์†์— ๊ฒฝ์Ÿ์‚ฌ ๋Œ€๋น„ ์ฐจ๋ณ„์  ๊ฒฝ์Ÿ์šฐ์œ„ & ์ง„์ž…์žฅ๋ฒฝ์ด๋ผ๋Š” ์ปจ์…‰์ด ๋ณ„๋กœ ์—†๋‹ค. (์ ์–ด๋„ MBA ์ˆ˜์—…์— ์˜ค์…จ๋˜ ๋ถ„๋“ค์˜ ์ด์•ผ๊ธฐ๋ฅผ ์ญ‰ ๋“ค์–ด๋ดค์„ ๋•Œ ๊ทธ๋ ‡๊ณ , ํ˜„์ง€์—์„œ ๋Œ€ํ™”ํ•˜๋Š” ๋ถ„๋“ค์˜ ์ด์•ผ๊ธฐ ๊ฒฝ์ฒญ ์‹œ ๊ทธ๋ ‡๋‹ค).

๊ทธ๋“ค์˜ ๋งˆ์Œ์†์— ์žˆ๋Š” ๋‘ ๊ฐ€์ง€ ํ‚ค์›Œ๋“œ๋Š”, ์œ ์ €, ๊ทธ๋ฆฌ๊ณ  ํŒ€์ด๋‹ค. ์œ ์ €์˜ ๋งŒ์กฑ์„ ์œ„ํ•ด Day 1์˜ ๋งˆ์Œ์œผ๋กœ ๋” ์•…์ฐฉ๊ฐ™์ด ์ตœ์„ ์„ ๋‹คํ•˜๋Š” ๊ฒƒ, ๊ทธ๋ฆฌ๊ณ  Day 1์˜ ๋งˆ์Œ์œผ๋กœ ์›€์ง์ด๋Š” ํŒ€์„ ๋”๋”๋” ์˜ค๋ž˜ ์œ ์ง€ํ•˜๋Š” ๊ฒƒ์ด, ์œ„๋Œ€ํ•œ ๊ธฐ์—…์„ ์ผ๊ตฐ ์ฐฝ์—…๊ฐ€๋“ค์ด ๊ณตํ†ต์ ์œผ๋กœ ๊ฐ•์กฐํ–ˆ๋˜ ๋ถ€๋ถ„์ด๋‹ค. (์šฐ๋ฒ„๋Š” ์ดˆ๋ฐ˜์— ์ง„์ž…์žฅ๋ฒฝ์ด ์žˆ์—ˆ์„๊นŒ? ์—์–ด๋น„์—”๋น„๋‚˜ ์œ„๋Œ€ํ•œ B2B SaaS ํšŒ์‚ฌ๋“ค์€ ๊ทธ๋Ÿฌํ–ˆ์„๊นŒ? ๊ทธ๋“ค์ด ๋งŒ๋“  scale ์ด ์ง„์ž…์žฅ๋ฒฝ์ผ ์ˆ˜ ์žˆ๋Š”๋ฐ, ๊ทธ scale ์€ IP ๊ฐ€ ์•„๋‹Œ ์ง‘์ฐฉ๊ณผ ๋…ธ๋ ฅ์˜ ์‹œ๊ฐ„์ด ๋งŒ๋“ค์–ด ์ค€ ์‚ฐ๋ฌผ์ด ์•„๋‹๊นŒ? ์œ ์ €์— ๋Œ€ํ•œ ์ง‘์ฐฉ์ด ์œ ์ €๋ฅผ ์œ„ํ•œ ์„œ๋น„์Šค/๊ธฐ์ˆ ์„ ๋งŒ๋“ค์–ด ๋‚ด๊ณ , ๊ทธ ๊ธฐ์ˆ /๊ธฐ๋Šฅ์ด ํ”ํžˆ ์„ธ์ƒ์ด ํ‰๊ฐ€ํ•˜๋Š” ๋…์ ์  ๊ธฐ์ˆ ์ด์ง€๋Š” ์•Š์„๊นŒ? ๊ทธ๋ ‡๋‹ค๋ฉด ๊ทธ๋“ค์˜ ์ง„์ž…์žฅ๋ฒฝ์€ ๊ฒฐ๊ณผ๋กœ ๋งŒ๋“ค์–ด์ง„ ๋…์ ์  ๊ธฐ์ˆ ์ผ๊นŒ? ์•„๋‹ˆ๋ฉด ๊ทธ ๊ธฐ์ˆ ์„ ๋งŒ๋“ค์–ด ๋‚ธ ํŒ€์˜ ์ง‘์š”ํ•จ์ผ๊นŒ? ํ•œ ๋•Œ ํ›Œ๋ฅญํ•œ ๊ธฐ์—…์ด, ํŒ€์ด ๋ฌด๋„ˆ์ง€๋Š” ๊ณผ์ •์—์„œ ํ•œ ์ˆœ๊ฐ„์— ๋ฌด๋„ˆ์ง€๋Š” ๊ฒƒ์„ ๋ณด๋ฉด, market position ์„ ์ง€์ผœ์ฃผ๋Š” ์˜์›ํ•œ ๊ธฐ์ˆ ์ด ์กด์žฌํ•˜๋Š” ๊ฒƒ์ด ๋งž๋Š”๊ฐ€? ๊ทธ๋ž˜์„œ ์•„๋งˆ์กด์€ Day 1 ์„ ๊ฐ•์กฐํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹๊นŒ?)

ํ˜„๋Œ€ ์‚ฌํšŒ์—์„œ๋Š” ์šฐ๋ฆฌ ์„œ๋น„์Šค๋ฅผ ์ข‹์•„ํ•ด ์ฃผ๋Š” ์œ ์ €๊ฐ€ ๋งŽ์•„์ง€๋Š” ๊ฒƒ์ด ์ง„์ž…์žฅ๋ฒฝ์ด๊ณ  ์›์ฒœ๊ธฐ์ˆ ์ด๋‹ค. ์œ ์ €๊ฐ€ ์šฐ๋ฆฌ ์„œ๋น„์Šค๋ฅผ ์ข‹์•„ํ•ด์ฃผ๋Š” ์ด์œ ๋Š”, ์šฐ๋ฆฌ ํšŒ์‚ฌ๊ฐ€ ๋…์ ์  ๊ธฐ์ˆ , ์ฐจ๋ณ„์  ๊ฒฝ์Ÿ์šฐ์œ„๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ์–ด์„œ๋ผ๊ธฐ ๋ณด๋‹ค๋Š”... ๋‹ค์†Œ ๋ถ€์กฑํ•œ ์ ์ด ์žˆ์–ด๋„ (๋ฒ„๊ทธ๊ฐ€ ์žˆ๊ณ , ๋ถˆํŽธํ•จ์ด ์žˆ์–ด๋„) ์œ ์ €์—๊ฒŒ ์ตœ์„ ์„ ๋‹คํ•˜๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค.

์œ ์ €์—๊ฒŒ ์ตœ์„ ์„ ๋‹คํ•˜๋Š” ๊ฒƒ์ด ์–ด๋ ต๋‚˜? ๋˜๋ฌผ์œผ๋ฉด, "์ด ์„ธ์ƒ์— ๋ชจ๋“  ํŒ€์ด ์œ ์ €์—๊ฒŒ ๋ช‡ ๋…„ ๊ฐ„ ๊พธ์ค€ํžˆ ์ง„์‹ฌ์œผ๋กœ ์ตœ์„ ์„ ๋‹คํ•˜๋Š” ํšŒ์‚ฌ๊ฐ€ ์–ผ๋งˆ๋‚˜ ์žˆ์„๊นŒ์š”? ๊ทธ๋Ÿฐ ๊ฒฝํ—˜ ์ตœ๊ทผ์— ํ•˜์‹  ์  ์žˆ์œผ์‹ค๊นŒ์š”?" "์ง€๊ธˆ ๊ท€ํ•˜์˜ ํšŒ์‚ฌ ์ž„์ง์›์€ ์˜ค๋กœ์ง€ ์œ ์ €๋งŒ ๋ฐ”๋ผ๋ณด๋ฉฐ ํ•˜๋ฃจ๋ฅผ ์ง‘์ค‘ํ•˜๊ณ  ์žˆ๋‚˜์š”? ์˜คํžˆ๋ ค ๋ง์”€ํ•˜์‹œ๋Š” ๋ถ„ (์˜ˆ: ์ž„์›)์˜ ๋ฐ”๋ฅผ ๋„˜์„๊นŒ๋ฅผ ๋” ๊ณ ๋ฏผํ•˜๊ณ  ์žˆ์ง€๋Š” ์•Š์„๊นŒ์š”? ๊ทธ ๊ณผ์ •์—์„œ ๋” ์ข‹์€ ์ด์ง์˜ ๊ธฐํšŒ๋ฅผ ์—†์„๊นŒ ๊ณ ๋ฏผํ•˜๊ณ  ์žˆ๋Š” ๋ถ„์ด ์žˆ์ง€๋Š” ์•Š์„๊นŒ์š”?"

์š”์ฆ˜์€ ํŠนํ—ˆ๋กœ ๊ธฐ์ˆ  ์‚ฌ์šฉ์„ ๋ง‰๊ธฐ ๋ณด๋‹ค๋Š”, open API ๋กœ ๋ฐฐํฌํ•˜๋Š” ์‹œ๋Œ€์ด๋‹ค. ๊ฒฝ์Ÿ์‚ฌ๋Š” ์˜คํžˆ๋ ค ์„œ๋กœ ๋‹ค๋ฅธ ์†Œ๊ตฌ์ ์„ ๊ธฐ๋ฐ˜์œผ๋กœ ํ•จ๊ป˜ ์—…์„ ํ‚ค์›Œ๋‚˜๊ฐ€๋Š” ๋‚˜๊ฐ€๋Š” ๋™๋ฃŒ์ธ ์‹œ๋Œ€์ด๋‹ค. ์ด๋Ÿฐ ์‹œ๋Œ€์—์„œ์˜ ํ•ต์‹ฌ์—ญ๋Ÿ‰, ์ฐจ๋ณ„์  ๊ฒฝ์Ÿ์šฐ์œ„, ์ง„์ž…์žฅ๋ฒฝ์€, ํšŒ์‚ฌ๊ฐ€ ์ •์˜ํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹Œ ์œ ์ €๊ฐ€ ์ •์˜ํ•ด ์ฃผ์‹œ๊ณ  ์ด ์‚ฌํšŒ์— ์ •์˜ํ•ด ์ฃผ๋Š” ๊ฒƒ์ด๋‹ค. ๊ทธ๊ฒƒ์ด tangible ํ•œ๊ฐ€? ์˜ค๋ž˜ํ† ๋ก ์ง€์†๋  ์ˆ˜ ์žˆ๋Š”๊ฐ€? ๋ฌป๋Š”๋‹ค๋ฉด, "๊ทธ๋ž˜์„œ ํŒ€์ด ์ค‘์š”ํ•˜๊ณ , ๋ฌธํ™”๊ฐ€ ์ค‘์š”ํ•˜๊ณ , ๋งˆ์Œ๊ฐ€์ง์ด ์ค‘์š”ํ•˜๊ณ , ์‚ฌ๋žŒ์ด ์ค‘์š”์š”ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ ๋ฃจํ”„๊ฐ€ ๋ฌด๋„ˆ์ง€๋ฉด ๋งŽ์€ ๊ฒƒ์ด ๋ฌด๋„ˆ์ง‘๋‹ˆ๋‹ค" ๋ผ๊ณ  ๋ง์”€๋“œ๋ฆฌ๊ณ  ์‹ถ๋‹ค.

์‚ฌ๋žŒ์„ ์œ„ํ•œ ์„œ๋น„์Šค๋Š” ์‚ฌ๋žŒ์ด ๋งŒ๋“œ๋Š” ๊ฒƒ์ด๋‹ค. ๊ทธ๋ฆฌ๊ณ , ์‚ฌ๋žŒ์€ ์‚ฌ๋žŒ์„ ์•Œ์•„๋ณธ๋‹ค.

https://lnkd.in/gDB7Juxr
Continuous Learning_Startup & Investment
https://m.youtube.com/watch?v=CsruQYKISYI&feature=youtu.be
๋ถˆํ™•์‹ค์„ฑ ์†์—์„œ ๋น ๋ฅด๊ฒŒ ์›€์ง์ด๊ณ  ๋ฐฉํ–ฅ์„ ์ฐพ์•„ ๋‚˜์„œ๊ธฐ ์œ„ํ•œ 10๊ฐ€์ง€ ๋ฐฉ๋ฒ• by Jeremy (Co-founder of Rippling and Director of Product Management at Coinbase)
๐Ÿ‘‰(๊ด‘๊ณ ) Startup, Investment, Science, Life ๋“ฑ ๋‹ค์–‘ํ•œ ์ฃผ์ œ์— ๊ด€์‹ฌ์ด ์žˆ๋‹ค๋ฉด: https://t.me/+oonhLBMoVtdjNjI1

1. ํญ๋ณด๋‹ค๋Š” ๊นŠ์ด: ํ—จ๋ฆญ์Šจ์€ ๋ฆฌ๋”๊ฐ€ ํ”„๋ ˆ์ž„์›Œํฌ์— ์˜์กดํ•˜๊ธฐ๋ณด๋‹ค๋Š” ๋ฌธ์ œ๋ฅผ ๊นŠ์ด ํŒŒ๊ณ ๋“ค์–ด ํ•ด๋‹น ๋ถ„์•ผ์˜ ์ „๋ฌธ๊ฐ€๊ฐ€ ๋  ๊ฒƒ์„ ๊ถŒ์žฅํ–ˆ์Šต๋‹ˆ๋‹ค.
2. ๋„์ „์ ์ธ MVP ์‚ฌ๊ณ ๋ฐฉ์‹: ๊ทธ๋Š” ์ตœ์†Œ๊ธฐ๋Šฅ์ œํ’ˆ(MVP) ์ ‘๊ทผ ๋ฐฉ์‹์ด ์ฐฝ์˜์„ฑ์„ ์ œํ•œํ•˜๊ณ  ์ž˜๋ชป๋œ ๊ธฐ์ˆ ์  ๊ฒฐ์ •์„ ๋‚ด๋ฆด ์ˆ˜ ์žˆ๋‹ค๊ณ  ๊ฒฝ๊ณ ํ–ˆ์Šต๋‹ˆ๋‹ค. ๋‹น์žฅ ์ง€์›๋˜์ง€ ์•Š๋”๋ผ๋„ ๊ฐ€์žฅ ๋ณต์žกํ•œ ์‚ฌ์šฉ ์‚ฌ๋ก€๋ฅผ ๋จผ์ € ์„ค๊ณ„ํ•˜๋ฉด ํ–ฅํ›„ ํ™•์žฅ์„ฑ๊ณผ ์ ์‘์„ฑ์„ ํ™•๋ณดํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
3. ์†Œ๊ทœ๋ชจ ํŒ€์„ ์œ„ํ•œ ๋ช…ํ™•ํ•œ ๋ฏธ์…˜: ๊ทธ๋Š” ๋ช…ํ™•ํ•œ ๋ฏธ์…˜์„ ๊ฐ€์ง„ ์†Œ๊ทœ๋ชจ ํŒ€์€ ์ƒํ™ฉ์— ๋งž๋Š” ์˜์‚ฌ๊ฒฐ์ •์„ ๋‚ด๋ฆฌ๊ณ  ๊ทœ๋ชจ์— ๋งž๋Š” ์†๋„๋ฅผ ์œ ์ง€ํ•˜๋ฉด์„œ ๋” ๋น ๋ฅด๊ฒŒ ๋Œ€์‘ํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ์ ‘๊ทผ ๋ฐฉ์‹์€ ์ฝ”์ธ๋ฒ ์ด์Šค์—์„œ 40๋ฐฐ ์„ฑ์žฅํ•˜๋Š” ๋™์•ˆ ๊ทธ์˜ ์žฌ์ž„ ๊ธฐ๊ฐ„ ๋™์•ˆ ์ค‘์š”ํ•œ ์—ญํ• ์„ ํ–ˆ์Šต๋‹ˆ๋‹ค.
4. ํ˜„์žฅ ์šด์˜ ์ดํ•ด: ํ—จ๋ฆญ์Šจ์€ ๋ฌธ์ œ, ๊ณผ์ œ, ์„ฑ๊ณต์„ ์ดํ•ดํ•˜๊ธฐ ์œ„ํ•ด 'ํ˜„์žฅ์˜ ํŒ€'๊ณผ ์ง์ ‘ ์†Œํ†ตํ•˜์—ฌ ์ตœ๊ณ  ์ˆ˜์ค€์˜ ์ œํ’ˆ ๊ฒฐ์ •์ด ํ˜„์žฅ์˜ ํ˜„์‹ค๊ณผ ์ผ์น˜ํ•˜๋„๋ก ํ•˜๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•˜๋‹ค๋Š” ์‹ ๋…์„ ๊ณต์œ ํ–ˆ์Šต๋‹ˆ๋‹ค.
5. ๋น ๋ฅด๊ฒŒ ๋ณ€ํ™”ํ•˜๋Š” ๋ฌธํ™”์™€ ์˜์‚ฌ ๊ฒฐ์ •: ํ—จ๋ฆญ์Šจ์€ ๋ฆฌํ”Œ๋ง์˜ ๋ฌธํ™”๋ฅผ ๋น ๋ฅธ ์˜์‚ฌ ๊ฒฐ์ • ๋Šฅ๋ ฅ์ด ์š”๊ตฌ๋˜๋Š” '์†๋„'์˜ ๋ฌธํ™”๋ผ๊ณ  ์„ค๋ช…ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฅผ ์œ„ํ•ด์„œ๋Š” ์šฐ์„ ์ˆœ์œ„์™€ ๊ทธ๋ ‡์ง€ ์•Š์€ ์šฐ์„ ์ˆœ์œ„๋ฅผ ๋ช…ํ™•ํžˆ ํ•ด์•ผ ํ•˜๋ฉฐ, ์ด๋ฅผ ํ†ตํ•ด ๋ชจ๋“  ์‚ฌ๋žŒ์ด ์ž์‹ ์˜ ์—ญ๋Ÿ‰์„ ์ตœ๋Œ€ํ•œ ๋ฐœํœ˜ํ•  ์ˆ˜ ์žˆ์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
6. ์ „๋ฌธ์„ฑ๊ณผ ๋””ํ…Œ์ผ ์ค‘์‹ฌ์˜ ์ ‘๊ทผ ๋ฐฉ์‹: ๊ทธ๋Š” ์ œํ’ˆ ๋ฆฌ๋”๊ฐ€ ์ œํ’ˆ์˜ ์„ธ๋ถ€ ์‚ฌํ•ญ์— ๋Œ€ํ•œ ์„ธ๊ณ„์ ์ธ ์ „๋ฌธ๊ฐ€๊ฐ€ ๋˜์–ด ๋งŽ์€ ์ •๋ณด๋ฅผ ํก์ˆ˜ํ•˜๊ณ  ๋ถˆํ™•์‹ค์„ฑ์„ ํƒ์ƒ‰ํ•˜๋ฉฐ ์˜์‚ฌ ๊ฒฐ์ •์— ์ž์‹ ๊ฐ์„ ๊ฐ€์ ธ์•ผ ํ•œ๋‹ค๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค.
7. ํ•œ๋ฐœ ์•ž์„œ ์ƒ๊ฐํ•˜๊ธฐ: ๊ทธ๋Š” ๊ธ€๋กœ๋ฒŒ ํ™•์žฅ์„ ์œ„ํ•œ ๊ณ„ํš์„ ์ƒ๊ฐ๋ณด๋‹ค ์ผ์ฐ ์„ธ์šฐ๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•˜๋‹ค๊ณ  ๊ฐ•์กฐํ–ˆ์Šต๋‹ˆ๋‹ค. ๋ชจ๋“  ๊ตญ๊ฐ€๋Š” ๊ณ ์œ ํ•œ ํŠน์„ฑ์„ ๊ฐ€์ง€๊ณ  ์žˆ์œผ๋ฏ€๋กœ ๋ฏธ๊ตญ๊ณผ ๋™์ผํ•œ ์ ‘๊ทผ ๋ฐฉ์‹์„ ์ ์šฉํ•˜๋Š” ๊ฒƒ์€ ํšจ๊ณผ์ ์ด์ง€ ์•Š์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
8. ํ”„๋ ˆ์ž„์›Œํฌ์™€ ํ”„๋กœ์„ธ์Šค: ํ—จ๋ฆญ์Šจ์€ ํ”„๋ ˆ์ž„์›Œํฌ๊ฐ€ ๋„์›€์ด ๋  ์ˆ˜ ์žˆ์ง€๋งŒ ํŒ€์˜ ํŠน์ • ๋ผ์ดํ”„์‚ฌ์ดํด์— ๋งž๊ฒŒ ์กฐ์ •๋˜์–ด์•ผ ํ•œ๋‹ค๊ณ  ๋ฏฟ์Šต๋‹ˆ๋‹ค. ํ”„๋กœ์„ธ์Šค์— ์ง€๋‚˜์น˜๊ฒŒ ์˜์กดํ•˜๋ฉด ์ œํ’ˆ์— ๋Œ€ํ•œ ๊นŠ์€ ์‚ฌ๊ณ ๋ฅผ ๋ฐฉํ•ดํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
9. ์ฑ„์šฉ ์ฒ ํ•™: ํ—จ๋ฆญ์Šจ์€ ์ œํ’ˆ ๊ด€๋ฆฌ์ž๋ฅผ ์ฑ„์šฉํ•  ๋•Œ ์ •์‹ ์  ๋ฏผ์ฒฉ์„ฑ๊ณผ ํ†ต์ฐฐ๋ ฅ ์žˆ๋Š” ์งˆ๋ฌธ์„ ์ค‘์‹œํ•˜๋ฉฐ, ํฐ ๋น„์ฆˆ๋‹ˆ์Šค ๊ทธ๋ฆผ๊ณผ ์„ธ๋ถ€์ ์ธ ์งˆ๋ฌธ ๋ชจ๋‘์— ๋Œ€ํ•ด ์ƒ๊ฐํ•  ์ˆ˜ ์žˆ๋Š” ์ง€์›์ž์˜ ์ค‘์š”์„ฑ์„ ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. mental agility and insightful questioning, emphasizing the importance of candidates.
10. ๋น„์ฆˆ๋‹ˆ์Šค ์šฐ์„ ์ˆœ์œ„์™€ '๋‹จํ˜ธํ•จ(Imperatives)': ๋ฆฌํ”Œ๋ง์€ ๋‹ค์Œ ๋ถ„๊ธฐ ๋˜๋Š” 6๊ฐœ์›” ๋™์•ˆ ์šฐ์„ ์ˆœ์œ„๊ฐ€ ์ง€์ •๋œ ์—…๋ฌด ๋ชฉ๋ก์ธ 'ํ•„์ˆ˜ ๊ณผ์ œ'๋ฅผ ๋„์ž…ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด๋ฅผ ํ†ตํ•ด ๊ฐœ๋ณ„ ํŒ€ ๋ชฉํ‘œ์™€ ์ค‘์š”ํ•œ ํšŒ์‚ฌ ๋ชฉํ‘œ ์‚ฌ์ด์˜ ๊ท ํ˜•์„ ์œ ์ง€ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
This is a pretty cool idea, right?
Within the paper, the authors reveal the professions in which 100% of the work will be impacted by LLMs : mathematicians, tax preparation, financial analysts, writers, & web designers. Insurance appraisers, financial managers, & search marketing strategists will see less than 15% of their work impacted by AI.
What do you think? Will large-language models produce greater productivity gains than the personal computer?
https://tomtunguz.com/llm-impact-gdp/
The appeal of AI isnโ€™t just the technology.
Rather, potential astronomical revenue growth fuels valuations by bringing software to new categories & by serving new unmet interest & voracious demand.
The businesses capturing that demand will command massive valuation premiums.
Even with an AI moniker adorning the pitch deck, a software company is a software company.
https://www.linkedin.com/pulse/does-ai-premium-exist-fundraising-market-tomasz-tunguz[โ€ฆ]9uifaGFiA%253D%253D/?trackingId=dndFk7IhRimb19uifaGFiA%3D%3D
https://www.linkedin.com/pulse/what-every-saas-app-spoke-english-tomasz-tunguz%3FtrackingId=Mob7bZWUSbulj2X%252BlgpMfg%253D%253D/?trackingId=Mob7bZWUSbulj2X%2BlgpMfg%3D%3D

์ด๋ฒˆ ์ฃผ์— Hubspot์—์„œ ์ƒ์œ„ 10๋ช…์˜ ํ™•์žฅ ๋ฆฌ๋“œ๋ฅผ ์ฐพ์€ ๋‹ค์Œ, ๊ฐ ๋ฆฌ๋“œ์˜ ๋กœ๊ณ ๋ฅผ ์ˆ˜์ง‘ํ•˜๊ณ  Adobe Firefly์—์„œ ๊ฒฝ์ฃผ์šฉ ์ž๋™์ฐจ์— ํšŒ์‚ฌ ๋กœ๊ณ ๊ฐ€ ์žˆ๋Š” ์ด๋ฏธ์ง€๋ฅผ ๋งŒ๋“ค์–ด ์ด๋ฉ”์ผ์— ์ฒจ๋ถ€ ํŒŒ์ผ๋กœ ์ฒจ๋ถ€ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Ÿฐ ๋‹ค์Œ ๊ณ ๊ฐ ์ง€์› ํ‹ฐ์ผ“์˜ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๊ฐ ์ž ์žฌ ๊ณ ๊ฐ์—๊ฒŒ ์ด๋ฉ”์ผ ์ดˆ์•ˆ์„ ์ž‘์„ฑํ•˜๊ณ  ๋‚ด ์ดˆ์•ˆ ํด๋”์— ์ €์žฅํ•ฉ๋‹ˆ๋‹ค.

์ฒซ์งธ, ๋ฐ์ดํ„ฐ ๋ณด์•ˆ ๋ฐ ๋ฐ์ดํ„ฐ ์†์‹ค ๋ฐฉ์ง€์ž…๋‹ˆ๋‹ค. ๋ณด์•ˆ ์ฑ…์ž„์ž๋Š” ์›Œํฌํ”Œ๋กœ๊ฐ€ ์Šน์ธ๋œ ์‚ฌ๋žŒ์— ์˜ํ•ด, ํ—ˆ์šฉ๋œ ๋ฐ์ดํ„ฐ์— ๋Œ€ํ•ด, ์Šน์ธ๋œ ๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์—ฌ ์‹คํ–‰๋˜๊ณ , ๊ทธ ํ”„๋กœ์„ธ์Šค๊ฐ€ ๊ตญ์ œ ๋ฐ์ดํ„ฐ ๊ทœ์ •์„ ์ค€์ˆ˜ํ•˜๋Š”์ง€ ์–ด๋–ป๊ฒŒ ๋ณด์žฅํ•  ์ˆ˜ ์žˆ์„๊นŒ์š”?
๋‘˜์งธ, ์˜์–ด API(์ผ๋ช… LLM)๋Š” ํ™•๋ฅ ์ ์ž…๋‹ˆ๋‹ค. ์‚ฌ๋žŒ์ฒ˜๋Ÿผ ์‹ค์ˆ˜๋ฅผ ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ž ์žฌ์  ์˜ค๋ฅ˜๋Š” ์‹ฌ๊ฐํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค(๋ชจ๋“  Hubspot CRM ๋ ˆ์ฝ”๋“œ์˜ ์†Œ์œ ์ž๋ฅผ ํ˜„์žฌ ์‚ฌ์šฉ์ž๋กœ ์—…๋ฐ์ดํŠธํ•œ๋‹ค๊ณ  ์ƒ์ƒํ•ด ๋ณด์„ธ์š”). ๋ชจ๋‹ˆํ„ฐ๋ง, ํ…Œ์ŠคํŠธ ๋ฐ ๋กค๋ฐฑ/์‹คํ–‰ ์ทจ์†Œ ๋ฒ„ํŠผ์ด ๋งค์šฐ ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค.
์ง€๋‚œ ์ˆ˜์‹ญ ๋…„ ๋™์•ˆ ๊ฐœ๋ฐœ์ž๋“ค์€ ์ฝ”๋“œ ๋ฆฌ๋ทฐ, ํ…Œ์ŠคํŠธ, ๋ชจ๋‹ˆํ„ฐ๋ง, ๊ฒฉ๋ฆฌ ๋“ฑ ์ฝ”๋“œ๋ฅผ ์œ„ํ•ด ์ด์™€ ๊ฐ™์€ ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ•ด ์™”์Šต๋‹ˆ๋‹ค. ํšŒ์‚ฌ์˜ ๋ชจ๋“  ์‚ฌ๋žŒ์ด ์ž์—ฐ์–ด API๋ฅผ ํ†ตํ•ด ์†Œํ”„ํŠธ์›จ์–ด๋ฅผ ์˜ค์ผ€์ŠคํŠธ๋ ˆ์ด์…˜ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋œ๋‹ค๋ฉด ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜ ๊ฐœ๋ฐœ์„ ์œ„ํ•œ ๋™๋“ฑํ•œ ๋„๊ตฌ๋„ ํ•„์š”ํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
๋ชจ๋ธ ๋ฐ ์ œ์–ด๊ฐ€ ๊ฐœ์„ ๋จ์— ๋”ฐ๋ผ ์ž์—ฐ์–ด API๋ฅผ ํ†ตํ•œ SaaS ์˜ค์ผ€์ŠคํŠธ๋ ˆ์ด์…˜์€ ์ƒ์‚ฐ์„ฑ์„ ํฌ๊ฒŒ ํ–ฅ์ƒ์‹œํ‚ฌ ๊ฒƒ์ž…๋‹ˆ๋‹ค.
"Thanks to the astonishing growth in the capabilities of generative AI, we believe SaaS is now entering its fourth generation: a system of cognition."

https://medium.com/lightspeed-venture-partners/saas-4-0-say-hello-to-the-era-of-cognition-cb22d549b460
https://www.nytimes.com/2023/06/14/technology/europe-ai-regulation.html

The AI Act has several implications for AI startup founders and investors, particularly in the areas of compliance and regulatory challenges, impact on innovation and development, global influence and competitiveness, and investment opportunities and risks.

Compliance and regulatory challenges: The AI Act introduces a risk-based approach with different requirements for each level of risk. Startups developing AI systems will need to adhere to the regulatory framework, which may entail additional costs and efforts, particularly for high-risk AI systems. Businesses will need to invest in compliance management, internal audits, and reporting to ensure they meet the Act's requirements.

Impact on innovation and development: The AI Act encourages innovation in the development of AI systems that are more transparent, accountable, and aligned with societal values. Companies that can develop innovative solutions to address the challenges posed by the AI Act will have a significant competitive advantage and will be better positioned to capitalize on the growing AI market.

Global influence and competitiveness: The AI Act is likely to set a global standard, impacting businesses developing and using AI worldwide. As a result, companies like OpenAI, Google, and Microsoft may be required to declare whether copyrighted material has been used to train their AI systems. The legislation's impact is so significant that OpenAI, the maker of ChatGPT, has stated it may be forced to pull out of Europe depending on the final text.

Investment opportunities and risks: The AI Act creates new opportunities for collaboration between businesses, research institutions, and regulatory bodies. Companies that can successfully navigate the complex regulatory environment and collaborate with stakeholders will be better positioned to attract investment. However, some investors are concerned about the potential impact of the AI Act on the competitiveness of European AI startups. According to a survey conducted by a coalition of AI-focused institutions, 73% of venture capitalists expect the AI Act to reduce or significantly reduce the competitiveness of European startups in AI.

In summary, the AI Act presents both challenges and opportunities for AI startup founders and investors. Navigating the regulatory landscape and fostering innovation while adhering to the Act's requirements will be crucial for startups to succeed and attract investment in the evolving AI market.
์ฐฝ์—…์ž์˜ ๋ฒˆ์•„์›ƒ์€ SaaS์˜ ์ง„์ •ํ•œ ์œ„ํ—˜์ž…๋‹ˆ๋‹ค. ๋ณดํ†ต ์ดˆ๊ธฐ์—๋Š” ๊ทธ๋ ‡์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์‹œ๊ฐ„์ด ์ง€๋‚ ์ˆ˜๋ก ํฐ ์œ„ํ—˜์ž…๋‹ˆ๋‹ค. ์ข…์ข… ๋ชจ๋“  ๊ฒƒ์ด ๋งˆ์นจ๋‚ด ์™„์„ฑ๋˜๋Š” ๊ฒƒ์ฒ˜๋Ÿผ ๋ง์ด์ฃ .

์žฅ๊ธฐ์ ์ธ ๋ฒˆ์•„์›ƒ์„ ํ”ผํ•˜๊ธฐ ์œ„ํ•œ 10๊ฐ€์ง€ ์ œ์•ˆ์„ ์†Œ๊ฐœํ•ฉ๋‹ˆ๋‹ค:

1/ ์ง„์ •ํ•œ ์†Œ์œ ์ž์ธ ๋ถ€์‚ฌ์žฅ์„ ์ถ”๊ฐ€๋กœ ๊ณ ์šฉํ•˜์„ธ์š”. ์—ฌ๊ธฐ์„œ ๋ˆ ๋ช‡ ํ‘ผ ์•„๋ผ๋ ค๊ณ  ํ•˜์ง€ ๋งˆ์„ธ์š”.

์—ฌ๋Ÿฌ๋ถ„์€ ๋ชจ๋“  ๊ฒƒ์„ ์žˆ๋Š” ๊ทธ๋Œ€๋กœ ์ฑ…์ž„์ ธ์•ผ ํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทœ๋ชจ๋ฅผ ํ™•์žฅํ• ์ˆ˜๋ก ์†Œ์œ ํ•ด์•ผ ํ•  ๊ฒƒ์€ ์ค„์–ด๋“ญ๋‹ˆ๋‹ค. ์ž์‹ ์˜ ์—…๋ฌด ์˜์—ญ์„ ์ฑ…์ž„์งˆ ์ •๋ง ํ›Œ๋ฅญํ•œ ๋ถ€์‚ฌ์žฅ์„ ํ•œ ๋ช…, ๋‘ ๋ช…, ์„ธ ๋ช… ์ฐพ์•„๋ณด์„ธ์š”. ๊ทธ๋Ÿฌ๋ฉด ์ •์‹ ์ ์ธ ๋ถ€๋‹ด์„ ํฌ๊ฒŒ ์ค„์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

2/ COO ๊ณ ์šฉ

์ด๊ฒƒ์€ ๊ธฐ๋ฐœํ•˜๊ณ  ์ด์ƒํ•œ ๊ณ ์šฉ์ด์ง€๋งŒ, 5๋ฐฑ๋งŒ ๋‹ฌ๋Ÿฌ์˜ ARR์ด๋ผ๋„ ํ›Œ๋ฅญํ•œ ์‚ฌ๋žŒ์„ ์ฐพ์„ ์ˆ˜ ์žˆ๋‹ค๋ฉด ์–ด์จŒ๋“  ์‹œ๋„ํ•ด ๋ณด์„ธ์š”. ํšŒ์‚ฌ ์šด์˜์˜ 20 ~ 40 %๋ฅผ ์†Œ์œ  ํ•  ์ˆ˜์žˆ๋Š” ์‚ฌ๋žŒ์„ ์ฐพ์œผ์‹ญ์‹œ์˜ค.

3/ ๋น„์ฆˆ๋‹ˆ์Šค ํด๋ž˜์Šค๋กœ ๋น„ํ–‰ํ•˜๊ณ , ๊ดœ์ฐฎ์€ ํ˜ธํ…”์— ๋จธ๋ฌผ๊ธฐ

์šฐ๋ฆฌ ๋ชจ๋‘๊ฐ€ ์ด๊ฒƒ์„ ์›ํ•˜์ง€ ์•Š๋Š”๋‹ค๋Š” ๊ฒƒ์„ ์•Œ๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์—ฌํ–‰์—๋Š” ๋Œ€๊ฐ€๊ฐ€ ๋”ฐ๋ฆ…๋‹ˆ๋‹ค. CEO๊ฐ€ 3์ฃผ ๋™์•ˆ ์ด๋™ ์ค‘์—๋„ ํ—ˆ๋ฆฌ ํ†ต์ฆ์ด ์žˆ์–ด์„œ๋Š” ์•ˆ ๋ฉ๋‹ˆ๋‹ค. ๊ทธ๊ฒƒ์€ ๋„์›€์ด๋˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

4 / ํœด๊ฐ€ ๊ฐ€๊ธฐ

1๋…„์— 2์ฃผ ์ •๋„๋Š” ์–ด๋””๋ก ๊ฐ€ ๋– ๋‚˜์„ธ์š”. ํ•˜๋ฃจ์— ๋‘ ๋ฒˆ ์ด๋ฉ”์ผ์„ ํ™•์ธํ•˜์ง€๋งŒ ๊ทธ๊ฒŒ ๋‹ค์ž…๋‹ˆ๋‹ค. ์Šคํ…Œ์ด์ผ€์ด์…˜์€ ํ•˜์ง€ ๋งˆ์„ธ์š”.

5 / ์ง„์งœ ๋ฉ˜ํ†  ์ฐพ๊ธฐ

์ด๋Ÿฐ์ €๋Ÿฐ ์ด์•ผ๊ธฐ๋ฅผ ๋‚˜๋ˆŒ ์ˆ˜ ์žˆ๋Š” ์‚ฌ๋žŒ์ด ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค. ์ง„์งœ๋กœ. ๊ทธ๋“ค์—๊ฒŒ ๊ณตํ‰ํ•˜๊ฒŒ ์ง€๋ถˆํ•˜์‹ญ์‹œ์˜ค.

6/ ๋ช‡ ๋‹ฌ๋Ÿฌ ๋” ๋ชจ๊ธˆํ•˜๊ธฐ

๋ฒค์ฒ˜ ์ง€์›์„๋ฐ›๋Š” ๊ฒฝ์šฐ ๊ณ„ํš ํ•œ ๊ฒƒ๋ณด๋‹ค 20 % ๋” ๋งŽ์ด ๋ชจ์œผ์‹ญ์‹œ์˜ค. ์—ฐ์†Œ์œจ์„ ๋†’์ด๊ฑฐ๋‚˜ ์ง€์ถœํ•˜์ง€ ์•Š์œผ๋ฉด ์ด๊ฒƒ์€ ๋‹น์‹ ์˜ ์‚ถ์„ ๋งŽ์ด ๊ดด๋กญํž ๊ฒƒ์ž…๋‹ˆ๋‹ค.

7/ ์ฒซํ•ด ์ดํ›„์—๋Š” ์ง‘์— ๊ฐ€์ ธ๊ฐ€์ง€ ๋งˆ์„ธ์š”.

์ฒซํ•ด๊ฐ€ ์ง€๋‚˜๋ฉด ์Šคํƒ€ํŠธ ์—…์˜ ์ŠคํŠธ๋ ˆ์Šค์™€ ๋“œ๋ผ๋งˆ๊ฐ€ ๊ฐ€์กฑ์—๊ฒŒ ๋„ˆ๋ฌด ๋งŽ์„ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ดˆ์ฐฝ๊ธฐ ์ดํ›„์—๋Š” ๋„ˆ๋ฌด ๋งŽ์ด ๊ณต์œ ํ•˜์ง€ ๋งˆ์„ธ์š”. ์ง‘์€ ๊ฐ€์กฑ์„ ์œ„ํ•œ ๊ณต๊ฐ„์ด์ž ์ŠคํŠธ๋ ˆ์Šค๋ฅผ ํ’€ ์ˆ˜ ์žˆ๋Š” ๊ณณ์ด์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๊ฐ€์กฑ์€ ๋Š์ž„์—†๋Š” ๋“œ๋ผ๋งˆ๋ฅผ ๊ฒฌ๋”œ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค. ๊ฑฐ์˜ ๋ชจ๋“  ์‚ฌ๋žŒ์—๊ฒŒ ๋„ˆ๋ฌด ๋งŽ์€ ์ผ์ž…๋‹ˆ๋‹ค.

8/ ๋œ ํƒœ์šฐ๊ธฐ

๋ˆ ๋ถ€์กฑ์— ๋Œ€ํ•ด ๊ฑฑ์ •ํ•  ํ•„์š”๊ฐ€์—†๋Š” ์œ„์น˜์— ์ž์‹ ์„ ๋‘์‹ญ์‹œ์˜ค. ์ž์‹ ์˜ ์šด๋ช…์„ ํ†ต์ œํ•˜๋Š” ๊ฒƒ์€ ๋ฉ‹์ง„ ์ผ์ž…๋‹ˆ๋‹ค. ๋ฒค์ฒ˜ ๋ผ์šด๋“œ๋ฅผ ๋ชจ์„ ์ˆ˜ ์žˆ๋”๋ผ๋„ ๋ชจ๋‘ ์“ฐ์ง€ ๋งˆ์„ธ์š”. ํ•„์š” ์ด์ƒ์œผ๋กœ ํ•˜๋ฃจ๋ผ๋„ ๋” ์™ธ๋ถ€์— ์˜์กดํ•˜์—ฌ ์ƒ๋ช… ์œ ์ง€์— ์˜์กดํ•˜์ง€ ๋งˆ์„ธ์š”. ์ด๋Ÿฐ๋‹ค๊ณ  ํ•ด์„œ ๋ˆˆ์‚ด์ด ์ฐŒํ‘ธ๋ ค์ง€์ง€๋Š” ์•Š์„ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ทธ๊ฒƒ์€ ์—„์ฒญ๋‚˜๊ฒŒ ํ•ด๋ฐฉ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.

9/ ๋” ๋นจ๋ฆฌ ์„ฑ์žฅํ•˜๊ธฐ (๋•Œ๋•Œ๋กœ)

๋•Œ๋กœ๋Š” ๋” ๋นจ๋ฆฌ ์„ฑ์žฅํ•˜๋Š” ๊ฒƒ์ด ์ŠคํŠธ๋ ˆ์Šค๋ฅผ ํ›จ์”ฌ ๋œ ๋ฐ›์Šต๋‹ˆ๋‹ค. ๋„ˆ๋ฌด ๋А๋ฆฌ๊ฒŒ ์„ฑ์žฅํ•˜๋ฉด ์ž์›์ด ๋ถ€์กฑํ•ด์ง‘๋‹ˆ๋‹ค. ๋ชจ๋“  ๊ฒƒ์ด ๋А๋ ค์งˆ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ฑ„์šฉ์ด ๋” ์–ด๋ ค์›Œ์ง‘๋‹ˆ๋‹ค. ์–ธ๋ก  ํ™๋ณด๊ฐ€ ๋” ์–ด๋ ค์›Œ์ง‘๋‹ˆ๋‹ค. ๊ณ ๊ฐ๊ณผ ํŒŒํŠธ๋„ˆ์˜ ๊ด€์‹ฌ๋„ ๋” ์–ด๋ ค์›Œ์ง‘๋‹ˆ๋‹ค. ๋•Œ๋กœ๋Š” ๋А๋ฆฌ๊ฒŒ ์„ฑ์žฅํ•˜๋Š” ๊ฒƒ์ด ์ŠคํŠธ๋ ˆ์Šค๊ฐ€ ๋œํ•  ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์žฅ๊ธฐ์ ์œผ๋กœ๋Š” ๋” ํฐ ์ŠคํŠธ๋ ˆ์Šค๊ฐ€ ๋˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ์Šต๋‹ˆ๋‹ค.

10/ ํ–‰๋ณตํ•œ ๊ณ ๊ฐ์— ์ง‘์ค‘

๋งŽ์€ ๊ฒƒ๋“ค์ด ๋„ˆ๋ฌด ์–ด๋ ต์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋ˆ„๊ตฌ๋‚˜ ๊ณ ๊ฐ ํ–‰๋ณต์„ ํ–ฅ์ƒ์‹œํ‚ฌ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ๋‹ค๋ฅธ ์•„์ด๋””์–ด๊ฐ€ ์—†๋”๋ผ๋„ NPS์™€ CSAT๋ฅผ ํฌํ•จํ•˜์—ฌ์ด๋ฅผ ํ–ฅ์ƒ ์‹œํ‚ค์‹ญ์‹œ์˜ค. ์ด๊ฒƒ์€ ์žฅ๊ธฐ์ ์œผ๋กœ ๋‹น์‹ ์„ ๊ดด๋กญํž ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๋‹จ๊ธฐ์ ์œผ๋กœ๋Š” ๊ทธ๋ ‡๊ฒŒ ๋А๊ปด์ง€์ง€ ์•Š์„์ง€๋ผ๋„ ๋†’์€ ๊ณ ๊ฐ ์ดํƒˆ์€ ์žฅ๊ธฐ์ ์œผ๋กœ ๋งค์šฐ ์ŠคํŠธ๋ ˆ์Šค๋ฅผ ์ค๋‹ˆ๋‹ค. ๋ชจ๋“  ์Šคํƒ€ํŠธ์—…์€ NPS๋ฅผ ๋†’์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๊ฒƒ์€ ํ•ญ์ƒ ํšจ๊ณผ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.

๊ฒฐ๊ตญ 4~5๋…„๋งˆ๋‹ค ์ž์‹ ๊ณผ ํšŒ์‚ฌ๋ฅผ ์žฌ์ฐฝ์กฐํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋ ‡๊ฒŒ ํ•  ์ˆ˜ ์žˆ๋‹ค๋ฉด ์ตœ์†Œ 4~5๋…„์€ ๋” ๊ฐ€์ž…ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋ ‡์ง€ ์•Š๋‹ค๋ฉด. ๊ทธ ๋ฌด๋ ต๋ถ€ํ„ฐ ์„œ์„œํžˆ ๋ฌด๋„ˆ์ง€๊ธฐ ์‹œ์ž‘ํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. MRR์ด ๊ฝค ์˜ค๋žซ๋™์•ˆ ๊ณ„์† ์ฆ๊ฐ€ํ•˜๋”๋ผ๋„

https://www.linkedin.com/feed/update/urn:li:activity:7075090159125331970/
โค2
OpenAI has introduced several updates to its API, including new function calling capabilities, more steerable versions of GPT-4 and GPT-3.5 Turbo, and a 75% cost reduction in the Embeddings model

These updates have implications for builders who have integrated with the OpenAI API or ChatGPT Plugins, as well as new opportunities for startup founders.Improved function calling: The new function calling capability in the Chat Completions API allows developers to more reliably connect the power of GPT models with external tools1. This feature enables developers to get structured data back from the model, making it easier to create chatbots that answer questions by calling external tools, like ChatGPT Plugins2

Enhanced steerable models: The updated GPT-4 and GPT-3.5 Turbo models provide better control over the AIโ€™s behavior, enabling developers to build more sophisticated and customized AI applications1
Longer context and lower prices: The new 16k context version of GPT-3.5 Turbo allows developers to work with longer text inputs, while the 75% cost reduction in the Embeddings model makes it more affordable for developers to use the API1

These updates create new opportunities for startup founders to leverage OpenAIโ€™s technology in innovative ways. For example, startups can build AI-powered applications that take advantage of the improved function calling capabilities and steerable models to create more advanced and customized solutions1. Additionally, the cost reductions make it more accessible for startups to experiment with and integrate OpenAIโ€™s technology into their products3.

Furthermore, the Azure OpenAI Service, which provides access to OpenAIโ€™s large language models like GPT-3.5 and Codex, presents more opportunities for startups in Microsoftโ€™s Founders Hub to build using the next generation of AI services4. This access to cutting-edge AI technology can help startups develop new applications and solutions, attracting investment and gaining a competitive advantage in the market.