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
https://twitter.com/saranormous/status/1669490272940736513

3/ Mostly, real products need to be:
*consistent
*tested
*versioned
*observable
*up to date
*fast enough
*cheap enough
*deterministic in their outputs (if composed, consumed)
*able to be reasoned about by the developer (and the end user!)

4/ This is incredibly hard, and doesnโ€™t come out of the box with stochastic LLMs and ambiguous natural language user input. The initial successes weโ€™ve seen in genAI apps happen to be cases and pioneering products that flex or fix some of the requirements.
AI๋ช…๊ฐ€ ์—…์Šคํ…Œ์ด์ง€์˜ ๊ธ€์ž…๋‹ˆ๋‹ค. ์–ธ์–ด ๋ชจ๋ธ์— ๋Œ€ํ•œ ์ „์ฒด ํžˆ์Šคํ† ๋ฆฌ๋ฅผ ์ดํ•ดํ•˜๊ธฐ์— ์ข‹์€ ๊ธ€์ด๋„ค์š”.

๋‹ค๋งŒ ๋งˆ๋ฌด๋ฆฌ ๋ถ€๋ถ„์€ ์ €๋Š” ์กฐ๊ธˆ ์• ๋งคํ•œ ๋А๋‚Œ์ž…๋‹ˆ๋‹ค. โ€œ๋ชจ๋ธ ๊ตฌ๋…โ€ ํ‘œํ˜„์ด ๋‚˜์ด๋ธŒํ•˜๊ธฐ๋„ ํ•˜๊ณ , ์–‘์งˆ ๋ฐ์ดํ„ฐ๋Š” ์ง€๊ธˆ๋ฟ๋งŒ์ด ์•„๋‹ˆ๋ผ 20๋…„ ์ „์—๋„ ์ค‘์š”ํ–ˆ๊ธฐ ๋•Œ๋ฌธ์— ํฐ ์ž„ํŒฉํŠธ๋Š” ์—†๋„ค์š”.

์ €๋Š” ๊ฐœ๋ฐœ ์ธก๋ฉด์—์„œ 3๊ฐ€์ง€ ํŒ€์˜ ๋“ฑ์žฅ์„ ๊ธฐ๋Œ€ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋ฌผ๋ก  ๋›ฐ์–ด๋‚œ UX์™€ ๋ช…ํ™•ํ•œ ๋ฌธ์ œ ์„ค์ •์ด ๊ฐ€์žฅ ์ค‘์š”ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.

1) ํ™•์žฅ์„ฑ ์ธก๋ฉด์—์„œ ๊ฐ€์žฅ ํ•„์š”ํ•˜๊ณ  ์„ฑ๊ณตํ•  ๊ฒƒ์ด๋ผ ์˜ˆ์ƒํ•˜๋Š” ํŒ€์€ DevOps, ์ฆ‰ ์ „์ฒด์  ํŒŒ์ดํ”„๋ผ์ธ ๊ตฌ์ถ•์ด ๋ชจ๋“ˆ๋Ÿฌํ•˜๊ณ  ์• ์ž์ผํ•˜๊ฒŒ ์ด๋ค„์ง€๋Š” ํŒ€์ด๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. ๊ฒฐ๊ตญ ๋ชจ๋ธ์€ ์ •์ฐฉํ•˜์ง€ ์•Š๊ณ  ์—…๋ฐ์ดํŠธ๊ฐ€ ๊พธ์ค€ํžˆ ์ด๋ค„์ง€๋ฉฐ ๋ฐ˜๋ณต์ ์ธ ๋ฐฐํฌ๊ฐ€ ์ด๋ค„์งˆ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ด ๊ณผ์ •์—์„œ ๋น„์šฉ ํšจ์œจ์„ฑ๊ณผ ๊ฒฝ์Ÿ๋ ฅ ๋ชจ๋“  ๊ฒƒ์ด ๊ฒฐ์ •๋˜์ง€ ์•Š์„๊นŒ ์ƒ๊ฐ์ด ๋“ญ๋‹ˆ๋‹ค.

2) ๋‘๋ฒˆ์งธ๋Š” ๋ฐ์ดํ„ฐ์˜ ์ˆ˜์ง‘์„ ๊ณ ๋„ํ™”ํ•  ์ˆ˜ ์žˆ๋Š” ํŒ€์ž…๋‹ˆ๋‹ค. ์‹ค์ œ ๋ฐ์ดํ„ฐ ๋ ˆ์ด๋ธ”๋ง์€ ๋…ธ๋™ ์ง‘์•ฝ์ ์ž…๋‹ˆ๋‹ค. ๋ฌผ๋ก  self-supervised learning ๋“ฑ์˜ ๋ฐœ์ „์œผ๋กœ ๋ ˆ์ด๋ธ”๋ง์ด ์ ์  ์‰ฌ์›Œ์ง€๊ณ  ์žˆ์ง€๋งŒ ์•„์ง ์—ฌ์ „ํžˆ ๋‹ค์ˆ˜ ๋ถ„์•ผ์—๋Š” ๋ฐ์ดํ„ฐ๋Š” ๋…ธ๋™์ง‘์•ฝ์ ์ธ ์ˆ˜์ง‘์ด ์ด๋ค„์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด ๊ณผ์ •์—์„œ ํˆด์˜ UX๋‚˜ ์ธ์„ผํ‹ฐ๋ธŒ ๋ชจ๋ธ์ด ์ƒ๋‹นํžˆ ์ค‘์š”ํ•ด์งˆ ๊ฒƒ ๊ฐ™์€๋ฐ ์•„์ง ์ด ์‹œ์žฅ์˜ 1๋“ฑ์€ ์—†์–ด๋ณด์ด๋„ค์š”.

3) ๊ฒฐ๊ตญ์€ ์ธํ”„๋ผ์ž…๋‹ˆ๋‹ค. GPU/FPGA๋ฅผ ํšจ์œจ์ ์œผ๋กœ ์‚ฌ์šฉํ•˜๊ฒŒ ํ•ด์ค„ ์ˆ˜ ์žˆ๋Š” ํŒ€์ด ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค. training๋„, inference๋„ ๋ง์ด์ฃ . ์ด ๋ ˆ๋ฒจ์—์„œ๋Š” ํ•˜๋“œ์›จ์–ด/์†Œํ”„ํŠธ์›จ์–ด์—์„œ ์ตœ์ ํ™”ํ•  ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์ด ๋‹ค์–‘ํ•ฉ๋‹ˆ๋‹ค. ๋ฐ์ดํ„ฐ ์„ผํ„ฐ ๊ตฌ์ถ•, AI์šฉ ๋ฐ˜๋„์ฒด๋ถ€ํ„ฐ CUDA ์ตœ์ ํ™”๊นŒ์ง€ ์—ฌ๋Ÿฌ๋ชจ๋กœ ๊ธฐ๋Œ€๊ฐ€ ๋˜๋„ค์š”. ๊ทผ๋ฐ ์ด๋Ÿฐ ์ž‘์—…์€ ๋‹ค์ˆ˜์˜ ์ „๋ฌธ๊ฐ€์™€ ๋งŽ์€ ๋น„์šฉ์ด ํ•„์š”ํ•œ ๋ฌธ์ œ๋ผ ๊ธฐ์กด ๋Œ€๊ธฐ์—…์ด ์ž˜ํ•  ๊ฐ€๋Šฅ์„ฑ์ด ๋†’์€ ํŽธ์ด๋ผ ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. ๊ฐ€๋” ์Šˆํผ ์ฒœ์žฌ๊ฐ€ ๋‚˜ํƒ€๋‚˜ ์ด๋Ÿฐ ๋ฌธ์ œ๋ฅผ ๊ฐœ์„ ํ•˜๊ธฐ๋„ ํ•˜๋Š”๋ฐ, ๋” ๋งŽ์ด ๋“ฑ์žฅํ•˜๊ธฐ๋ฅผ ๊ธฐ๋Œ€ํ•ฉ๋‹ˆ๋‹ค :)

์ €๋Š” Public AI๊ฐ€ ๋”์šฑ ํฐ ์‹œ์žฅ์— ๋‚˜์˜ค๊ธฐ ์œ„ํ•ด์„œ๋Š” ๋ฐœ์ƒ ๊ฐ€๋Šฅํ•œ Worst Case์— ๋Œ€ํ•œ ๊ทœ์ œ์™€ ์ธ์‹์ด ์ค‘์š”ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. ์˜๋ฃŒ, ์žฌ๋ฌด, ๋ฒ•๋ฅ  ๋“ฑ ๊ณ ๋น„์šฉ ๋ฌธ์ œ๋Š” โ€œ์ฑ…์ž„๊ฐ’โ€์ด ํฌํ•จ๋œ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. ๋ชจ๋“  ์œ„ํ—˜์„ ์†Œ๋น„์ž์—๊ฒŒ ๋„˜๊ธฐ๋Š” ๊ฒƒ๋„ ํ•˜๋‚˜์˜ ๋ฐฉ๋ฒ•์ด๊ฒ ์ง€๋งŒ ๋Œ€๋‹ค์ˆ˜ AI๋Š” ํŠน์ • ์ฃผ์ฒด์— ์˜ํ•ด ๊ฐœ๋ฐœ์ด ์ด๋ค„์งˆ ๊ฒƒ์ด๋ผ๋ฉด ์ฑ…์ž„ ์†Œ์ œ ๋“ฑ์— ๋Œ€ํ•œ ๊ทœ์ œ ๋ฌธ์ œ๋Š” ๋”์šฑ ์ค‘์š”ํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์—ฐ์žฅ์„  ์ƒ์—์‚ฌ ์š”์ƒˆ๋Š” ์ž˜ ์•ˆ๋ณด์ด์ง€๋งŒ XAI(Explainable AI, ์„ค๋ช…๊ฐ€๋Šฅํ•œ AI) ํŒ€๋“ค์˜ ๋ถ€ํ™œ๋„ ๊ธฐ๋Œ€ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

https://www.linkedin.com/posts/subin-an-841975110_data-centric-ai-%EA%B4%80%EC%A0%90%EC%9C%BC%EB%A1%9C-%EC%9E%AC%ED%95%B4%EC%84%9D%ED%95%98%EB%8A%94-%EC%9E%90%EC%97%B0%EC%96%B8%EC%96%B4%EC%B2%98%EB%A6%AC-%EA%B8%B0%EB%B0%98-history-activity-7075383077157441537-p5fG?utm_source=share&utm_medium=member_ios
์ตœ๊ทผ ์ž‘์„ฑํ•œ ๋ธ”๋กœ๊ทธ ๊ธ€(Head of Sales์˜ ์˜จ๋ณด๋”ฉ ์ฒดํฌ๋ฆฌ์ŠคํŠธ)์—์„œ ํฌ์–ด์บ์ŠคํŠธ ๋ฐฉ๋ฒ•๋ก ์„ ์–ธ๊ธ‰ํ–ˆ๋Š”๋ฐ, ๊ด€๋ จํ•˜์—ฌ ์กฐ๊ธˆ ๋” ์ž์„ธํžˆ ์„ค๋ช…ํ•ด ๋‹ฌ๋ผ๋Š” ๋ฌธ์˜๊ฐ€ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.

์ด์ „ ํšŒ์‚ฌ์—์„œ๋Š” ์ฃผ๋กœ Enterprise Account Executive ๋“ฑ IC (Individual Contributor) ์—ญํ• ์„ ๋งก๋‹ค๊ฐ€ ์ƒˆ๋กœ์šด ํšŒ์‚ฌ์—์„œ ์ฒ˜์Œ์œผ๋กœ Head of Sales ์—ญํ• ์„ ๋งก๊ฒŒ ๋˜์—ˆ์„ ๋•Œ, ์ฒซ ํฌ์–ด์บ์ŠคํŠธ ๋ถ„์„์ด ๋ง‰์—ฐํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. Head of Sales ์—†์ด ๋ณธ์ธ์ด ์ด์‚ฌํšŒ์˜ ์š”์ฒญ์œผ๋กœ ์ฒซ ํฌ์–ด์บ์ŠคํŠธ ๋ถ„์„์„ ํ•˜๋Š” ์ฐฝ์—…์ž ๋ฐ CEO๊ป˜๋„ ๋„์›€์ด ๋  ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค.

ํ˜น์€ B2B/SaaS ํˆฌ์ž๋ฅผ ์ตœ๊ทผ์— ์ง‘ํ–‰ํ•˜์—ฌ ํšจ๊ณผ์ ์œผ๋กœ ์ด์‚ฌํšŒ๋ฅผ ์ง„ํ–‰ํ•˜๊ณ  ์‹ถ์€ VC๋ถ„๋“ค๊ป˜๋„ ๋„์›€์ด ๋˜๋ฆฌ๋ผ ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.

์ €๋Š” ์ง€๊ธˆ๋„ ์ผ์ฃผ์ผ์— ๋‘ ๋ฒˆ ์”ฉ ํฌ์–ด์บ์ŠคํŠธ ๋ฏธํŒ…์„ ์ง„ํ–‰ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ํ•œ๋ฒˆ์€ ์˜์—…ํŒ€ ๋ชจ๋‘์™€ ํ•จ๊ป˜, ๋‹ค๋ฅธ ํ•œ๋ฒˆ์€ ์˜์—…ํŒ€์˜ ๋ฆฌ๋”์‹ญ ๊ทธ๋ฃน๊ณผ ํ•จ๊ป˜. ์ด๋ ‡๊ฒŒ ์ค‘์š”๋„๊ฐ€ ๋†’์€ ํฌ์–ด์บ์ŠคํŠธ ๋ฏธํŒ…์ด๊ธฐ์— ์ œ๊ฐ€ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ๋Š” ๋ช‡ ๊ฐ€์ง€ ๋ฐฉ๋ฒ•๋ก ์„ ์†Œ๊ฐœํ•˜๋ ค ํ•ฉ๋‹ˆ๋‹ค. ์ฐธ๊ณ ๋กœ ์ €๋Š” ์—ฌ๊ธฐ์„œ ์†Œ๊ฐœํ•˜๋Š” ๋ฐฉ๋ฒ•๋ก ์—์„œ ํ•œ ๋‘ ๊ฐ€์ง€๋ฅผ ์„ ํƒํ•˜๋Š” ๊ฒƒ์ด ๋‹Œ, ์ด ๋ฐฉ๋ฒ•๋ก ์„ ํ˜ผํ•ฉํ•˜์—ฌ ๋ถ„๊ธฐ ๋ณ„ ์˜ˆ์ธก์น˜์˜ ๋ฒ”์œ„๋ฅผ ๋ฝ‘์•„๋‚ด๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๊ด€๋ จํ•˜์—ฌ์„œ๋Š” ์ด ๊ธ€์˜ ํ›„๋ฐ˜๋ถ€์— ์ถ”๊ฐ€๋กœ ์„ค๋ช…ํ•˜๊ฒ ์Šต๋‹ˆ๋‹ค.

https://kimchihill.com/2023/06/18/kr-a-comprehensive-approach-to-sales-forecasting-for-b2b-saas-startups/
โค2
โค3
We just published my interview with Scott Belsky. It was one of my favorite conversations of the year.

A few highlights:

The virtues of slow shipping. The prevailing dogma is that startups should ship and iterate as quickly as possible. Scott believes thatโ€™s often counterproductive. You need to โ€œsurprise and delightโ€ your customers to create a product that grows organically. You canโ€™t do that by simply meeting a userโ€™s expectations; you must surpass them. Doing so takes time and polishing.

Innovating within a big company. When Adobe acquired his startup Behance for $150 million in 2012, Belsky struggled to see his long-term future at a big corporation. Eleven years later and the former entrepreneur is Adobeโ€™s Chief Product Officer. Working at the $227 billion software behemoth, Scott discovered he was a mission-driven entrepreneur rather than a serial one. Rather than restricting his entrepreneurial urges, Adobe has given Scott a platform to tackle them from a new vantage.

Investing in a great mentor. Scott Belsky counts entrepreneur and author Seth Godin as a friend and mentor. Once a year, Scott visits Godin in upstate New York to discuss his plans, relying on his host to pick them apart. This productive inquisition helps Scott allocate his time and effort more thoughtfully. Investing in this relationship has proven fruitful, both personally and professionally.

The coming age of hyper-personalization. We live in a โ€œgeneralizedโ€ world, per Scott. When we visit an online store or look at a menu, we see the full range of options. The Adobe executive expects future generations to have more personalized product experiences that filter out unsuitable or unwanted options. They will also have far less choice, as a result.

The overstated AI threat. Reports of AIโ€™s dangers are exaggerated, according to Scott. Though he expects new models will increase the sophistication of scams and other frauds, he doesnโ€™t foresee humans being made subservient to an almighty algorithm. Instead, we should worry about the fragility of modern society, under threat from war, dictatorships, and corruption.

https://thegeneralist.substack.com/p/scott-belsky
Scott Belsky's book "The Messy Middle" offers a wealth of stories and insights for founders, despite its rambling and sometimes repetitive style. Here are some key takeaways from the book for founders:

- Business success is forged in the difficult middle of any venture.
- Give yourself a much-needed sense of a short-term reward by manufacturing milestones.
- Itโ€™s crucial to stay self-aware during times of both triumph and hardship.
- The messy middle is not only about getting through the inevitable lows and setbacks that hit you mid-project. Muddling through the messy middle also means using the high points of your journey to accelerate your progress toward the finish line.
- To leverage these highs, you must recognize and build on the aspects of your venture which are working well.
- The last mile is a different sport. It may involve acquisition by a large corporation, which calls for understanding dealmaking via advisors and coaches.
- A great founder isnโ€™t necessarily a great finisher.
- It's important to be transparent about the middle stage of your startup, including when you have to let someone off, lay someone off, or kill a product that's not growing fast enough.
- Maximizing creative output and using constraints to boost productivity can be helpful in sparking creativity.
- Focus on the most important tasks to avoid getting bogged down in minutiae.
- Building a strong team, staying focused on the long-term vision, and being willing to take risks are essential for success.
"The future cannot be predicted, but futures can be invented" -- Dennis Gabor 1963.
"The best way to predict the future is to invent it" -- Alan Kay 1971

If you can't predict whether a technology is going to be beneficial or not, build it so it is.

https://lnkd.in/eHA23npu

Yann LeCun
๐Ÿ‘1
Continuous Learning_Startup & Investment
https://youtu.be/5cQXjboJwg0
The speaker believes that there's investment potential at both ends of the technology stack - at the silicon level (to prevent any single company from monopolizing the market) and at the top where unique data sets can drive learning. Middle-level opportunities may be more volatile, potentially worth billions one day and nothing the next.

Chamath Palihapitiya gave an interesting example involving the travel industry, specifically the relationship between travel aggregators (like Expedia) and data providers (like Sabre Corporation, denoted as SABR in the stock market). This example provides valuable lessons about building a moat as a startup founder.

Sabre Corporation is a travel technology company known for its Global Distribution System (GDS), which is used by travel agencies and companies to search for and book airline flights, hotels, and other travel-related services. Sabre's GDS holds extensive data from airlines, hotels, and other travel providers, which it normalizes and makes accessible to travel aggregators like Expedia.

Despite their smaller market capitalization, Sabre and similar data providers play an essential role in the travel industry. Without them, the larger travel aggregators wouldn't have access to the comprehensive and real-time data needed to operate effectively.

Chamath suggests that in the age of AI and machine learning, where a conversational interface could allow users to book travel arrangements within a chat or social media platform, the role of travel aggregators like Expedia could diminish. In this scenario, the real value lies in the data that companies like Sabre provide, as they are the linchpin that enables these AI-driven experiences.

Here are a few lessons for startup founders:

Data is King: Data, especially when unique and comprehensive, can be a significant moat. As Chamath noted, Sabre's value lies in its vast repository of data, which it can leverage even if the methods of delivering services to consumers change significantly.

Platform Neutrality: Sabre services a wide array of customers in the travel industry. By maintaining neutrality and providing essential services to all players, Sabre has ensured its continued relevance and demand.

Understand Where Value is Created: Knowing where value is created and preserved in your industry is crucial. Travel aggregators might have a higher market cap, but companies like Sabre have a moat because they control the data, an indispensable asset in the digital age.

Predict Industry Evolution: The travel industry, like many others, is being transformed by AI and machine learning. Understanding how these technologies can reshape your industry will allow you to position your startup correctly.

Pivot or Diversify if Necessary: If you find your company in a position similar to that of the travel aggregators, consider how you might pivot or diversify your offerings to retain value as the industry evolves. For instance, could you form partnerships or develop your own unique data sources?

Remember that even as you build your startup's moat, you should remain aware of the evolving technology landscape to ensure your moat remains relevant.
https://youtu.be/5cQXjboJwg0

๋ฐœํ‘œ์ž๋Š” ๊ธฐ์ˆ  ์Šคํƒ์˜ ์–‘์ชฝ ๋, ์ฆ‰ ์‹ค๋ฆฌ์ฝ˜ ์ˆ˜์ค€(ํŠน์ • ๊ธฐ์—…์ด ์‹œ์žฅ์„ ๋…์ ํ•˜๋Š” ๊ฒƒ์„ ๋ฐฉ์ง€ํ•˜๊ธฐ ์œ„ํ•œ)๊ณผ ๊ณ ์œ ํ•œ ๋ฐ์ดํ„ฐ ์„ธํŠธ๊ฐ€ ํ•™์Šต์„ ์ฃผ๋„ํ•  ์ˆ˜ ์žˆ๋Š” ์ตœ์ƒ์œ„ ์ˆ˜์ค€ ๋ชจ๋‘์— ํˆฌ์ž ์ž ์žฌ๋ ฅ์ด ์žˆ๋‹ค๊ณ  ๋ฏฟ์Šต๋‹ˆ๋‹ค. ์ค‘๊ฐ„ ์ˆ˜์ค€์˜ ๊ธฐํšŒ๋Š” ๋ณ€๋™์„ฑ์ด ๋” ์ปค์„œ ํ•˜๋ฃจ๋Š” ์ˆ˜์‹ญ์–ต ๋‹ฌ๋Ÿฌ์˜ ๊ฐ€์น˜๊ฐ€ ์žˆ๋‹ค๊ฐ€ ๋‹ค์Œ ๋‚ ์—๋Š” ์•„๋ฌด๊ฒƒ๋„ ์•„๋‹ ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค.

์ฐจ๋งˆ์Šค ํŒ”๋ฆฌํ•˜ํ”ผํ‹ฐ์•ผ๋Š” ์—ฌํ–‰ ์‚ฐ์—…, ํŠนํžˆ ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ(์˜ˆ: ์ต์Šคํ”ผ๋””์•„)์™€ ๋ฐ์ดํ„ฐ ์ œ๊ณต์—…์ฒด(์˜ˆ: ์„ธ์ด๋ฒ„ ์ฝ”ํผ๋ ˆ์ด์…˜, ์ฃผ์‹ ์‹œ์žฅ์—์„œ๋Š” SABR๋กœ ํ‘œ์‹œ๋จ)์˜ ๊ด€๊ณ„์— ๊ด€ํ•œ ํฅ๋ฏธ๋กœ์šด ์˜ˆ๋ฅผ ๋“ค์—ˆ์Šต๋‹ˆ๋‹ค. ์ด ์‚ฌ๋ก€๋Š” ์Šคํƒ€ํŠธ์—… ์ฐฝ์—…์ž๋กœ์„œ ํ•ด์ž ๊ตฌ์ถ•์— ๋Œ€ํ•œ ๊ท€์ค‘ํ•œ ๊ตํ›ˆ์„ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค.

์„ธ์ด๋ฒ„ ์ฝ”ํผ๋ ˆ์ด์…˜์€ ์—ฌํ–‰์‚ฌ ๋ฐ ๊ธฐ์—…์ด ํ•ญ๊ณต๊ถŒ, ํ˜ธํ…” ๋ฐ ๊ธฐํƒ€ ์—ฌํ–‰ ๊ด€๋ จ ์„œ๋น„์Šค๋ฅผ ๊ฒ€์ƒ‰ํ•˜๊ณ  ์˜ˆ์•ฝํ•˜๋Š” ๋ฐ ์‚ฌ์šฉํ•˜๋Š” ๊ธ€๋กœ๋ฒŒ ์œ ํ†ต ์‹œ์Šคํ…œ(GDS)์œผ๋กœ ์œ ๋ช…ํ•œ ์—ฌํ–‰ ๊ธฐ์ˆ  ํšŒ์‚ฌ์ž…๋‹ˆ๋‹ค. Sabre์˜ GDS๋Š” ํ•ญ๊ณต์‚ฌ, ํ˜ธํ…” ๋ฐ ๊ธฐํƒ€ ์—ฌํ–‰ ์ œ๊ณต์—…์ฒด์˜ ๋ฐฉ๋Œ€ํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ๋ณด์œ ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ, ์ด๋ฅผ ์ •๊ทœํ™”ํ•˜์—ฌ Expedia์™€ ๊ฐ™์€ ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ์—์„œ ์•ก์„ธ์Šคํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•ฉ๋‹ˆ๋‹ค.

์‹œ๊ฐ€์ด์•ก์€ ์ž‘์ง€๋งŒ Sabre์™€ ์œ ์‚ฌํ•œ ๋ฐ์ดํ„ฐ ์ œ๊ณต์—…์ฒด๋Š” ์—ฌํ–‰ ์—…๊ณ„์—์„œ ํ•„์ˆ˜์ ์ธ ์—ญํ• ์„ ๋‹ด๋‹นํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋“ค์ด ์—†๋‹ค๋ฉด ๋Œ€ํ˜• ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ๋“ค์€ ํšจ๊ณผ์ ์ธ ์šด์˜์— ํ•„์š”ํ•œ ์ข…ํ•ฉ์ ์ธ ์‹ค์‹œ๊ฐ„ ๋ฐ์ดํ„ฐ์— ์•ก์„ธ์Šคํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.

์ฐจ๋งˆ์Šค๋Š” ๋Œ€ํ™”ํ˜• ์ธํ„ฐํŽ˜์ด์Šค๋ฅผ ํ†ตํ•ด ์‚ฌ์šฉ์ž๊ฐ€ ์ฑ„ํŒ…์ด๋‚˜ ์†Œ์…œ ๋ฏธ๋””์–ด ํ”Œ๋žซํผ ๋‚ด์—์„œ ์—ฌํ–‰ ์ผ์ •์„ ์˜ˆ์•ฝํ•  ์ˆ˜ ์žˆ๋Š” AI ๋ฐ ๋จธ์‹ ๋Ÿฌ๋‹ ์‹œ๋Œ€๊ฐ€ ์˜ค๋ฉด ์ต์Šคํ”ผ๋””์•„์™€ ๊ฐ™์€ ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ๋“ค์˜ ์—ญํ• ์ด ์ค„์–ด๋“ค ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ์ด ์‹œ๋‚˜๋ฆฌ์˜ค์—์„œ ์ง„์ •ํ•œ ๊ฐ€์น˜๋Š” ์„ธ์ด๋ฒ„์™€ ๊ฐ™์€ ๊ธฐ์—…์ด ์ œ๊ณตํ•˜๋Š” ๋ฐ์ดํ„ฐ์— ์žˆ์œผ๋ฉฐ, ์ด๋Ÿฌํ•œ ๋ฐ์ดํ„ฐ๋Š” ์ด๋Ÿฌํ•œ AI ๊ธฐ๋ฐ˜ ๊ฒฝํ—˜์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜๋Š” ํ•ต์‹ฌ ์š”์†Œ์ด๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.


๋‹ค์Œ์€ ์Šคํƒ€ํŠธ์—… ์ฐฝ์—…์ž๋ฅผ ์œ„ํ•œ ๋ช‡ ๊ฐ€์ง€ ๊ตํ›ˆ์ž…๋‹ˆ๋‹ค:

๋ฐ์ดํ„ฐ๋Š” ์™•์ด๋‹ค: ๋ฐ์ดํ„ฐ, ํŠนํžˆ ๊ณ ์œ ํ•˜๊ณ  ํฌ๊ด„์ ์ธ ๋ฐ์ดํ„ฐ๋Š” ์ค‘์š”ํ•œ ํ•ด์ž๊ฐ€ ๋  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ฐจ๋งˆ์Šค๊ฐ€ ์–ธ๊ธ‰ํ–ˆ๋“ฏ์ด ์„ธ์ด๋ฒ„์˜ ๊ฐ€์น˜๋Š” ๋ฐฉ๋Œ€ํ•œ ๋ฐ์ดํ„ฐ ์ €์žฅ์†Œ์— ์žˆ์œผ๋ฉฐ, ์†Œ๋น„์ž์—๊ฒŒ ์„œ๋น„์Šค๋ฅผ ์ œ๊ณตํ•˜๋Š” ๋ฐฉ์‹์ด ํฌ๊ฒŒ ๋ณ€ํ™”ํ•˜๋”๋ผ๋„ ์ด๋ฅผ ํ™œ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

ํ”Œ๋žซํผ ์ค‘๋ฆฝ์„ฑ: ์„ธ์ด๋ฒ„๋Š” ์—ฌํ–‰์—…๊ณ„์˜ ๋‹ค์–‘ํ•œ ๊ณ ๊ฐ์—๊ฒŒ ์„œ๋น„์Šค๋ฅผ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค. ์ค‘๋ฆฝ์„ฑ์„ ์œ ์ง€ํ•˜๊ณ  ๋ชจ๋“  ํ”Œ๋ ˆ์ด์–ด์—๊ฒŒ ํ•„์ˆ˜์ ์ธ ์„œ๋น„์Šค๋ฅผ ์ œ๊ณตํ•จ์œผ๋กœ์จ ์„ธ์ด๋ฒ„๋Š” ์ง€์†์ ์ธ ๊ด€๋ จ์„ฑ๊ณผ ์ˆ˜์š”๋ฅผ ํ™•๋ณดํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.

๊ฐ€์น˜๊ฐ€ ์ฐฝ์ถœ๋˜๋Š” ๊ณณ์„ ์ดํ•ดํ•ฉ๋‹ˆ๋‹ค: ์—…๊ณ„์—์„œ ๊ฐ€์น˜๊ฐ€ ์ฐฝ์ถœ๋˜๊ณ  ๋ณด์กด๋˜๋Š” ์œ„์น˜๋ฅผ ํŒŒ์•…ํ•˜๋Š” ๊ฒƒ์€ ๋งค์šฐ ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค. ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ๋“ค์˜ ์‹œ๊ฐ€์ด์•ก์€ ๋” ๋†’์„ ์ˆ˜ ์žˆ์ง€๋งŒ, ์„ธ์ด๋ฒ„์™€ ๊ฐ™์€ ๊ธฐ์—…์€ ๋””์ง€ํ„ธ ์‹œ๋Œ€์— ์—†์–ด์„œ๋Š” ์•ˆ ๋  ์ž์‚ฐ์ธ ๋ฐ์ดํ„ฐ๋ฅผ ํ†ต์ œํ•˜๊ธฐ ๋•Œ๋ฌธ์— ํ•ด์ž๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

์—…๊ณ„์˜ ์ง„ํ™”๋ฅผ ์˜ˆ์ธกํ•ฉ๋‹ˆ๋‹ค: ๋‹ค๋ฅธ ๋งŽ์€ ์‚ฐ์—…๊ณผ ๋งˆ์ฐฌ๊ฐ€์ง€๋กœ ์—ฌํ–‰ ์‚ฐ์—…๋„ AI์™€ ๋จธ์‹ ๋Ÿฌ๋‹์œผ๋กœ ๋ณ€ํ™”ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ด๋Ÿฌํ•œ ๊ธฐ์ˆ ์ด ์—…๊ณ„๋ฅผ ์–ด๋–ป๊ฒŒ ์žฌํŽธํ•  ์ˆ˜ ์žˆ๋Š”์ง€ ์ดํ•ดํ•˜๋ฉด ์Šคํƒ€ํŠธ์—…์˜ ํฌ์ง€์…”๋‹์„ ์˜ฌ๋ฐ”๋ฅด๊ฒŒ ์„ค์ •ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

ํ•„์š”ํ•œ ๊ฒฝ์šฐ ํ”ผ๋ฒ— ๋˜๋Š” ๋‹ค๊ฐํ™”: ๊ท€์‚ฌ๊ฐ€ ์—ฌํ–‰ ์• ๊ทธ๋ฆฌ๊ฒŒ์ดํ„ฐ์™€ ๋น„์Šทํ•œ ์œ„์น˜์— ์žˆ๋‹ค๋ฉด, ์—…๊ณ„๊ฐ€ ์ง„ํ™”ํ•จ์— ๋”ฐ๋ผ ๊ฐ€์น˜๋ฅผ ์œ ์ง€ํ•˜๊ธฐ ์œ„ํ•ด ์ œํ’ˆ์„ ํ”ผ๋ฒ—ํ•˜๊ฑฐ๋‚˜ ๋‹ค๊ฐํ™”ํ•  ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์„ ๊ณ ๋ คํ•˜์„ธ์š”. ์˜ˆ๋ฅผ ๋“ค์–ด, ํŒŒํŠธ๋„ˆ์‹ญ์„ ๋งบ๊ฑฐ๋‚˜ ์ž์ฒด ์ƒํ’ˆ์„ ๊ฐœ๋ฐœํ•  ์ˆ˜ ์žˆ์„๊นŒ์š”?
์–ผ๋งˆ์ „์— ํŒŒํŠธํƒ€์ž„ CTO๋กœ ์ผํ•˜๊ณ  ์žˆ๋Š” ์Šคํƒ€ํŠธ์—… ํšŒ์‚ฌ์—์„œ ์ฃผ๋‹ˆ์–ด ๊ฐœ๋ฐœ์ž๋“ค๊ณผ ๋Œ€ํ™” ์ค‘.

CTO: ์šฐ๋ฆฌ๊ฐ€ ๊ฐœ๋ฐœํ•˜๊ณ  ์žˆ๋Š” ์„œ๋น„์Šค์˜ ์†Œ์Šค ์ฝ”๋“œ๋Š” ๊ธฐ์ˆ  ์ž์‚ฐ์ธ๊ฐ€์š”? ์•„๋‹ˆ๋ฉด ์‚ฌ์—… ์ž์‚ฐ์ธ๊ฐ€์š”?

๊ฐœ๋ฐœ์ž: ๊ธฐ์ˆ  ์ž์‚ฐ์ด ์•„๋‹๊นŒ์š”? ๊ฐœ๋ฐœ์˜ ๊ฒฐ๊ณผ๋ฌผ์ด๋‹ˆ๊นŒ์š”.

CTO: ๋งŒ์•ฝ ์‚ฌ์—… ๋ถ€์„œ๊ฐ€ ๊ฐœ๋ฐœํ•œ ์„œ๋น„์Šค๊ฐ€ ์‚ฌ์—…์ ์œผ๋กœ ์‹คํŒจํ•œ๋‹ค๋ฉด ์†Œ์Šค ์ฝ”๋“œ๋Š” ์–ด๋–ป๊ฒŒ ๋ ๊นŒ์š”? ์žฌํ™œ์šฉ ๊ฐ€๋Šฅํ• ๊นŒ์š”?

๊ฐœ๋ฐœ์ž: ์Œ... ๊ฑฐ์˜ ์žฌํ™œ์šฉ๋˜์ง€ ์•Š์„ ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค.

CTO: ๊ทธ๋Ÿฌ๋ฉด ์‹œ๊ฐ„์ด ์ง€๋‚˜๋ฉด ์•„๋ฌด๋„ ๊ด€์‹ฌ์„ ๊ฐ€์ง€์ง€ ์•Š์•„์„œ ์†Œ์Šค ์ฝ”๋“œ๊ฐ€ ์‚ฌ๋ผ์ง€๊ฒ ๋„ค์š”?

๊ฐœ๋ฐœ์ž: ์•„๋งˆ ๊ทธ๋ ‡๊ฒŒ ๋  ๊ฒƒ ๊ฐ™์Šต๋‹ˆ๋‹ค.

CTO: ๊ธฐ์ˆ  ์ž์‚ฐ์ธ๋ฐ ์‚ฌ์—… ๋ถ€์„œ๊ฐ€ ์‚ฌ์—…์„ ์ค‘๋‹จํ•˜๋ฉด ์‚ฌ๋ผ์ง„๋‹ค? ๋ญ”๊ฐ€ ์ด์ƒํ•˜์ง€ ์•Š๋‚˜์š”?

๊ฐœ๋ฐœ์ž: ์Œ...

CTO: ์‚ฌ์—… ๋ถ€์„œ๊ฐ€ ์–ดํ”Œ๋ฆฌ์ผ€์ด์…˜ ์†Œ์Šค ์ฝ”๋“œ์˜ ์šด๋ช…์„ ๊ฒฐ์ •ํ•œ๋‹ค๋ฉด ๊ทธ๊ฑด 100% ๊ธฐ์ˆ  ์ž์‚ฐ์œผ๋กœ ๋ณด๊ธฐ ํž˜๋“ญ๋‹ˆ๋‹ค. ์˜คํžˆ๋ ค ์‚ฌ์—… ์ž์‚ฐ์˜ ์„ฑ๊ฒฉ์ด ๊ฐ•ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ๊ทธ๋ ‡๋‹ค๋ฉด ๊ธฐ์ˆ  ์ž์‚ฐ์€ ๋ฌด์—‡์ธ๊ฐ€์š”?

CTO: ๊ธฐ์ˆ  ๋ถ€์ฑ„๋ผ๋Š” ์šฉ์–ด๋ฅผ ๋“ค์–ด๋ดค๋‚˜์š”?

๊ฐœ๋ฐœ์ž: ๋„ค.

CTO: ๊ธฐ์ˆ  ์ž์‚ฐ๊ณผ ๊ธฐ์ˆ  ๋ถ€์ฑ„๋Š” ์ข…์ด ํ•œ ์žฅ ์ฐจ์ด์ž…๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ๋ฐ˜๋Œ€ ์˜๋ฏธ ์•„๋‹Œ๊ฐ€์š”?

CTO: ์–ด๋–ค ์†Œ์Šค ์ฝ”๋“œ๊ฐ€ ๊ธฐ์ˆ  ์ž์‚ฐ์ด ๋˜๊ฑฐ๋‚˜ ๊ธฐ์ˆ  ๋ถ€์ฑ„๊ฐ€ ๋˜๋Š” ๊ธฐ์ค€์€ ๋ฌด์—‡์ผ๊นŒ์š”?

๊ฐœ๋ฐœ์ž: ๊น”๋”ํ•˜๊ฒŒ ์ž‘์„ฑ๋˜์–ด ์žˆ๋‹ค๋ฉด ์ž์‚ฐ์ด ๋˜๊ณ , ์ŠคํŒŒ๊ฒŒํ‹ฐ ์ฝ”๋“œ์ฒ˜๋Ÿผ ์žฌ์‚ฌ์šฉํ•˜๊ธฐ ํž˜๋“ค๊ฒŒ ๋ณต์žกํ•œ ์†Œ์Šค๋Š” ๋ถ€์ฑ„๊ฐ€ ๋˜๋Š” ๊ฒƒ์ด ์•„๋‹๊นŒ์š”?

CTO: ๊ทธ ๋ง๋„ ๋งž์•„์š”. ํ•˜์ง€๋งŒ ์ €๋Š” ๊ธฐ์ˆ  ์ž์‚ฐ๊ณผ ๊ธฐ์ˆ  ๋ถ€์ฑ„๋ฅผ ๋‚˜๋ˆ„๋Š” ๊ธฐ์ค€์€ ์‹œ์žฅ ๋‚ด์˜ ๊ฒฝ์Ÿ ๊ด€๊ณ„์— ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ๋ฌด์Šจ ๋œป์ธ๊ฐ€์š”?

CTO: ๊ฒฝ์Ÿ์ž๋ณด๋‹ค ๋” ์ €๋ ดํ•˜๊ณ  ๋น ๋ฅด๊ฒŒ ์ œํ’ˆ์„ ๊ฐœ๋ฐœํ•˜๊ฑฐ๋‚˜ ํ’ˆ์งˆ์ด ์šฐ์ˆ˜ํ•ด์ง€๋Š”๋ฐ ๊ธฐ์—ฌํ•œ๋‹ค๋ฉด ํ•ด๋‹น ์†Œ์Šค ์ฝ”๋“œ๋Š” ๊ธฐ์ˆ  ์ž์‚ฐ์ด ๋ฉ๋‹ˆ๋‹ค. ๋ฐ˜๋Œ€๋กœ ๊ฒฝ์Ÿ์ž๋ณด๋‹ค ๋น„์šฉ์ด ๋” ๋งŽ์ด ๋“ค๊ฑฐ๋‚˜ ์ œํ’ˆ ์ƒ์‚ฐ ์†๋„๊ฐ€ ๋А๋ฆฌ๋ฉฐ ์•ˆ์ •์„ฑ๋„ ๋–จ์–ด์ง„๋‹ค๋ฉด ๊ธฐ์ˆ  ๋ถ€์ฑ„๊ฐ€ ๋ฉ๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ์•„...

CTO: ๊ทธ๋ž˜์„œ ๋™์ผํ•œ ์†Œ์Šค ์ฝ”๋“œ๋ผ๋„ ์ฒ˜์Œ์—๋Š” ๊ธฐ์ˆ  ์ž์‚ฐ์ด์—ˆ์ง€๋งŒ ์‹œ์žฅ ํ™˜๊ฒฝ์˜ ๋ณ€ํ™”๋‚˜ ๊ธฐ์ˆ ์˜ ๋ฐœ์ „์œผ๋กœ ์ธํ•ด ๊ธฐ์ˆ  ๋ถ€์ฑ„๊ฐ€ ๋  ์ˆ˜๋„ ์žˆ์Šต๋‹ˆ๋‹ค. ํŠนํžˆ ์ œํ’ˆ์ด ๋ณต์žกํ•ด์ง€๋ฉด ์ฒ˜์Œ์—๋Š” ์ƒ๊ฐํ•˜์ง€ ๋ชปํ–ˆ๋˜ ๋‹จ์ˆœํ•œ ๊ตฌ์กฐ๋กœ ๊ฐœ๋ฐœ๋œ ์†Œ์Šค ์ฝ”๋“œ๊ฐ€ ๋‚˜์ค‘์— ๋ณต์žกํ•œ ๊ตฌ์กฐ์— ์ ์‘ํ•˜์ง€ ๋ชปํ•  ๋•Œ๋„ ๊ธฐ์ˆ  ๋ถ€์ฑ„๋กœ ๊ฐ„์ฃผ๋ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ทธ ๋ณธ์งˆ์€ ๊ฒฝ์Ÿ ๊ด€๊ณ„์˜ ๋น„๊ต ์šฐ์œ„์—์„œ ๊ธฐ์ˆ  ์ž์‚ฐ๊ณผ ๋ถ€์ฑ„๊ฐ€ ๊ฒฐ์ •๋œ๋‹ค๊ณ  ๋ด…๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ์–ด๋ ต๋„ค์š”.

CTO: ๋‹ค์‹œ ๊ธฐ์ˆ  ์ž์‚ฐ์€ ์–ด๋–ค ๊ฒƒ์ธ์ง€ ์งˆ๋ฌธ์œผ๋กœ ๋Œ์•„๊ฐ€๋ด…์‹œ๋‹ค. ๊ธฐ์ˆ  ์ž์‚ฐ์€ ๊ฒฝ์Ÿ์ž๋ณด๋‹ค ๋” ์ €๋ ดํ•˜๊ณ  ๋น ๋ฅด๊ฒŒ ๋†’์€ ํ’ˆ์งˆ์˜ ์ œํ’ˆ์„ ์ƒ์‚ฐํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋„์™€์ฃผ๋Š” ๊ฒƒ๋“ค๋กœ ์ •์˜ํ–ˆ์Šต๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์ด๋ฏธ ์ด๋Ÿฐ ๋…ธ๋ ฅ์„ ํ•˜๊ณ  ์žˆ์–ด์š”. ์˜ˆ๋ฅผ ๋“ค์–ด ๋‹ค์–‘ํ•œ ๊ฐœ๋ฐœ ๋ฌธ์„œ ์ž‘์—…, ๊ฐœ๋ฐœ ํ”„๋กœ์„ธ์Šค ์ •๋ฆฝ, ์ฝ”๋“œ ๋ฆฌ๋ทฐ, ๊ณตํ†ต ๋ชจ๋“ˆ์˜ ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌํ™” ์ž‘์—…, ํ”„๋กœ์ ํŠธ ์™„๋ฃŒ ํ›„ ์ฝ”๋“œ ์ตœ์ ํ™” ์ž‘์—… ๋“ฑ์„ ํ†ตํ•ด ๊ธฐ์ˆ  ์ž์‚ฐ์„ ํ™•๋ณดํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: CTO๋‹˜์ด ํ•ญ์ƒ ๊ฐ•์กฐํ•˜์‹œ๋Š” ๋‚ด์šฉ์ด๋„ค์š”. ์šฐ๋ฆฌ๊ฐ€ ๋ณ„๋กœ ์ข‹์•„ํ•˜์ง€ ์•Š๋Š” ์ž‘์—…๋“ค์ด์ง€๋งŒ์š”. ใ…Žใ…Ž

CTO: ใ…Žใ…Ž ๋งž์•„์š”. ๊ทธ๋ ‡์ง€๋งŒ ์ค‘์š”ํ•œ ์ž‘์—…๋“ค์ด์—์š”.

๊ฐœ๋ฐœ์ž: ๋ช‡ ๊ฐœ์›” ์ „๋ถ€ํ„ฐ CTO๋‹˜์ด ์ง€์‹œํ•˜์‹  ์šฐ๋ฆฌ ํšŒ์‚ฌ์˜ ๋ฆฌ์•กํŠธ UI ์ปดํฌ๋„ŒํŠธ๋“ค์˜ ํ‘œ์ค€ํ™” ๋ชจ๋“ˆ ์ž‘์—…๋„ ์ด์™€ ๊ด€๋ จ๋œ ์ผ์ด ๋งž๋‚˜์š”?

CTO: ๋„ค, ๊ทธ๊ฒŒ ๋ฐ”๋กœ ๊ทธ ์ผ์ž…๋‹ˆ๋‹ค. ๊ทธ ์ž‘์—…์€ ๋””์ž์ด๋„ˆ, ํผ๋ธ”๋ฆฌ์…”, ํ”„๋ก ํŠธ ์—”๋“œ ๊ฐœ๋ฐœ์ž๊ฐ€ ๋ชจ๋‘ ํšจ์œจ์ ์œผ๋กœ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ๋Š” UI ๋ชจ๋“ˆ์„ ํ‘œ์ค€ํ™”ํ•˜์—ฌ ๋น„์šฉ๊ณผ ์‹œ๊ฐ„์„ ์ ˆ์•ฝํ•  ์ˆ˜ ์žˆ๊ฒŒ ๋„์™€์ค๋‹ˆ๋‹ค. NPM์„ ํ†ตํ•œ ํ‘œ์ค€ ๋ชจ๋“ˆํ™” ์ž‘์—…์€ ์šฐ๋ฆฌ์˜ ๊ธฐ์ˆ  ์ž์‚ฐ์„ ํ™•๋ณดํ•˜๋Š” ์ค‘์š”ํ•œ ์˜ˆ์‹œ์ž…๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ์•„, ๊ทธ๋ ‡๊ตฐ์š”. ๊ทธ๋ž˜์„œ CTO๋‹˜์ด ๋‹น์žฅ ์‚ฌ์šฉ๋˜์ง€ ์•Š์„ ์š”์†Œ๊นŒ์ง€ ๊นŒ๋‹ค๋กญ๊ฒŒ ์š”๊ตฌํ•˜์…จ๋˜ ๊ฑฐ์˜€๊ตฐ์š”. ์–ธ์ œ ์‚ฌ์šฉ๋ ์ง€ ๋ชจ๋ฅด๋Š” ๊ฒƒ์„ ์™œ ๊ทธ๋ ‡๊ฒŒ ๊ฐ•์กฐํ•˜์‹œ๋Š”์ง€ ๊ถ๊ธˆํ–ˆ์–ด์š”.

CTO: ์–ดํ”Œ๋ฆฌ์ผ€์ด์…˜์—์„œ ์‚ฌ์šฉ๋˜๋Š” ์š”์†Œ๋“ค์€ ํ•„์š”ํ•  ๋•Œ๋งŒ ๊ฐœ๋ฐœํ•˜๊ณ  ์‚ฌ์šฉํ•˜๋ฉด ๋˜์ง€๋งŒ ํ‘œ์ค€ ๋ชจ๋“ˆ์ด๋‚˜ ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ ๊ฐœ๋ฐœ์€ ๋‹ค๋ฅธ ๋ฐฉ์‹์œผ๋กœ ์ ‘๊ทผํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ์˜๋„๋ฅผ ์ข€ ๋” ๋ถ„๋ช…ํ•˜๊ฒŒ ์ดํ•ดํ–ˆ์Šต๋‹ˆ๋‹ค.

CTO: ๊ฒฐ๋ก ์ ์œผ๋กœ ์šฐ๋ฆฌ ๊ฐœ๋ฐœ ์กฐ์ง์€ ๊ธฐ์ˆ  ๋ถ€์ฑ„๋ฅผ ์ค„์ด๊ณ  ๊ธฐ์ˆ  ์ž์‚ฐ์„ ๋Š˜๋ฆฌ๋Š” ๊ฒƒ์„ ์ž„๋ฌด๋กœ ํ•ฉ๋‹ˆ๋‹ค. ๊ธฐ์ˆ ์ ์œผ๋กœ ์„ ๋„ํ•˜๋Š” ํšŒ์‚ฌ๋ฅผ ๋งŒ๋“œ๋Š” ๊ฒƒ์€ ํฐ ๊ธฐ์ˆ  ๊ฐœ๋ฐœ ์ „๋žต์„ ์ˆ˜๋ฆฝํ•˜์—ฌ ๋‹ฌ์„ฑํ•˜๋Š” ๋ฐฉ๋ฒ•๋„ ์žˆ๊ณ  ๊ฐ์ž ๋งก์€ ์—…๋ฌด ๋‚ด์—์„œ ์†Œ์†Œํ•˜์ง€๋งŒ ์˜๋ฏธ ์žˆ๋Š” ๋…ธ๋ ฅ๋“ค์„ ํ†ตํ•ด ๊ธฐ์ˆ  ์ž์‚ฐ์„ ํ™•๋ณดํ•˜๋Š” ๊ฒƒ๋„ ๋˜ํ•œ ๊ฐ€๋Šฅํ•ฉ๋‹ˆ๋‹ค.

๊ฐœ๋ฐœ์ž: ์šฐ๋ฆฌ ์—ฐ๊ตฌ๊ฐœ๋ฐœ ์กฐ์ง์ด ํ•˜๋Š” ์ผ์„ ์ •ํ™•ํžˆ ์ดํ•ดํ–ˆ์Šต๋‹ˆ๋‹ค.

https://www.facebook.com/1054470436/posts/pfbid02D1VbTLPY3qHkaUgeA3MVrAQUMf2zxPLFjGQbEMAAFvfPByTuBoH1oHH2eWN6e5HJl/?mibextid=jf9HGS
โค2
Sources about languages in the internet

English is the most common language online, used by 60.4% of the top 10 million websites3. Other popular languages include Russian (8.5%), Spanish (4.0%), Turkish (3.7%), Persian (3.0%), French (2.6%), German (2.4%), and Japanese (2.1%) 3. As of January 2023, the share of websites using these languages is as follows: English (58.8%), Russian (5.3%), Spanish (4.3%), and French (2.9%) 4.

Based on the number of internet users, the top 10 languages and their market share on the internet are as follows1.

:English: 25.9%
Chinese: 19.4%
Spanish: 8.0%
Arabic: 5.2%
Portuguese: 4.1%
Indonesian/Malaysian: 3.7%
French: 3.5%
Japanese: 3.0%
Russian: 2.9%
German: 2.3%

These percentages represent the share of worldwide internet users who speak each language.On the other hand, based on the share of websites using these languages, the top 10 languages and their market share on the internet are as follows2.

:English: 58.8%
Russian: 5.3%
Spanish: 4.3%
French: 2.9%
German: 2.4%
Japanese: 2.1%
Portuguese: 2.1%
Italian: 1.9%
Persian: 1.8%
Chinese: 1.4%

Korean is an important language on the internet, with 39.4 million Korean-speaking internet users, representing 2% of the worldโ€™s total1. In 2010, Korean was ranked tenth among the languages used on the internet1.
โค1
1. **Market Size and Potential Growth for 10 years**:
The current global travel market is estimated at around $2.3 trillion annually in 2023, with significant components being airline revenue ($800B+) and global hotel demand ($1.5T). Over the next decade, this market is expected to grow significantly due to factors such as increasing global middle class, technological advancements, changing consumer preferences for experiences over material possessions, and the expected recovery and growth post the COVID-19 pandemic**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**.

2. **Major Players and Their Dynamics**:
The major players in travel distribution are split into several categories: Service Providers, Discovery engines, Pipes, and Metasearch companies**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**. The dynamics among these players often involve competitive and cooperative interactions, with changing power dynamics influenced by factors like technological innovation, consumer behavior, regulatory changes, and global events**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**.

3. **Current User Experience**:
Today's travel booking experience can often be fragmented and confusing**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**. This is mainly due to outdated technology, the multiplicity of service providers, and the inability of current systems to effectively bundle and sell ancillaries and upgrades**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**.

4. **Addressing the Issue - Scenarios, Winners, and Losers**:
- *Scenario 1: GDS-led change* โ€“ If GDS players like Sabre, Amadeus, and Travelport manage to modernize their systems and partner effectively with AI-based systems, they could continue to dominate.

[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)

- *Scenario 2: OTA-led change* โ€“ If OTAs manage to adapt, focusing on customer service and unique inventory, they could maintain their relevance.

[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)

- *Scenario 3: AI/LLM-led disruption* โ€“ If AI systems become advanced enough to make bookings directly on any platform, this would disrupt the entire industry.

[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)

5. **Opportunities for Startups**:

Startups can leverage AI to improve personalization, automate customer service, offer dynamic pricing, and disrupt current models**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**. A startup could develop an AI-powered concierge that streamlines the booking process, handles customer service, and personalizes travel recommendations**[2](https://chat.openai.com/?model=gpt-4#user-content-fn-2%5E)**.

6. **Investment Opportunities**:

Investors should consider investing in startups leveraging AI to disrupt the travel industry, particularly those focusing on improving customer experience, personalization, and service integration**[1](https://chat.openai.com/?model=gpt-4#user-content-fn-1%5E)**.
https://youtu.be/5cQXjboJwg0?t=4511

Brad Gerstner discusses the recent $105 million seed round raised by Mishro AI, a startup that has not yet written any code. He explains that rounds like this are necessary because these startups need to buy expensive hardware like h100s and a100s for training. However, Gerstner warns against making โ€œteaser bets,โ€ saying that this is not the way to make money in the industry. He compares the current AI funding frenzy to the search engine hype in the late 90s, where many companies went to zero despite the massive investment. Gerstner cautions against making too many โ€œstupid betsโ€ and stresses the importance of proper sizing and accountability within partnerships.

Brad Gerstner discusses how AI startups are overheating and how he sees history repeating itself, similar to the social networking market. He mentions a huge anti-portfolio for AI, saying no to over 60 companies, while many competitors are doing these deals, possibly just teaser bets. He warns that putting a bunch of small bets in order to buy a network, relationships or logos wonโ€™t work any better this time than it worked around social networking. He is concerned about LPs of venture funds, losing money, asking how the GPS are letting these things happen.

Brad Gerstner, the CEO of Altimeter Capital Management, shares his thoughts on overfunding AI startups. He advises that investing a large amount of money, such as $100 million, to buy compute for startups is not buying innovative IP but subsidizing capex and leading to a low-yield return. Gerstner explains that it is essential to invest in startups that are developing groundbreaking IP as it offers actual ownership and differentiation of the company. He emphasizes that when 80 cents/dollar goes towards sales and marketing, the returns are less. Gerstner adds that one should understand that putting $100 million into a startup is similar to leasing a function and that startups should look towards banking avenues such as Comerica Bank to fund their needs.

Brad Gerstner and Chamath Palihapitiya discuss the frenzy of funding in the AI startup industry and how itโ€™s creating financial illiteracy. While Gerstner believes a lot of good businesses are being funded, Palihapitiya states that pouring billions of dollars into something can lead to both product and distribution challenges.

the challenges of investing in AI startups and the importance of understanding how quickly model development and training is accelerating and how cost curves are moving. With cost reductions in machine learning happening much faster than Mooreโ€™s Law, the business model of AI startups should have some advantage in being the first to market, generating new data, and building a persisting advantage. However, itโ€™s difficult to invest in technology as itโ€™s unclear when the market advantage will lie compared to when technology tips, and the minimizing cost of inputs impacts the coordination cost as well. Moreover, he stated that itโ€™s easier to innovate and disrupt at the application and tools layers versus competing down the stack, where it takes more significant capital and might lead to zero returns.
๋ธŒ๋ž˜๋“œ ๊ฑฐ์ŠคํŠธ๋„ˆ๊ฐ€ ์•„์ง ์ฝ”๋“œ๋ฅผ ์ž‘์„ฑํ•˜์ง€ ์•Š์€ ์Šคํƒ€ํŠธ์—…์ธ ๋ฏธ์Šˆ๋กœ AI๊ฐ€ ์ตœ๊ทผ 1์–ต 5,500๋งŒ ๋‹ฌ๋Ÿฌ์˜ ์‹œ๋“œ ๋ผ์šด๋“œ๋ฅผ ์œ ์น˜ํ•œ ๊ฒƒ์— ๋Œ€ํ•ด ์ด์•ผ๊ธฐํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์ด๋Ÿฌํ•œ ์Šคํƒ€ํŠธ์—…์ด ํ›ˆ๋ จ์„ ์œ„ํ•ด h100 ๋ฐ a100๊ณผ ๊ฐ™์€ ๊ณ ๊ฐ€์˜ ํ•˜๋“œ์›จ์–ด๋ฅผ ๊ตฌ์ž…ํ•ด์•ผ ํ•˜๊ธฐ ๋•Œ๋ฌธ์— ์ด์™€ ๊ฐ™์€ ๋ผ์šด๋“œ๊ฐ€ ํ•„์š”ํ•˜๋‹ค๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ฒŒ๋ฅด์ŠคํŠธ๋„ˆ๋Š” โ€œํ‹ฐ์ € ๋ฒ ํŒ…โ€œ์„ ํ•˜๋Š” ๊ฒƒ์€ ์—…๊ณ„์—์„œ ๋ˆ์„ ๋ฒ„๋Š” ๋ฐฉ๋ฒ•์ด ์•„๋‹ˆ๋ผ๋ฉฐ ๊ฒฝ๊ณ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ํ˜„์žฌ์˜ AI ํŽ€๋”ฉ ์—ดํ’์„ 90๋…„๋Œ€ ํ›„๋ฐ˜์˜ ๊ฒ€์ƒ‰ ์—”์ง„ ๊ณผ๋Œ€๊ด‘๊ณ ์™€ ๋น„๊ตํ•˜๋ฉฐ, ๋งŽ์€ ๊ธฐ์—…์ด ๋ง‰๋Œ€ํ•œ ํˆฌ์ž์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  ์ ์ž๋ฅผ ๊ธฐ๋กํ–ˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ฑฐ์ŠคํŠธ๋„ˆ๋Š” ๋„ˆ๋ฌด ๋งŽ์€ โ€˜์–ด๋ฆฌ์„์€ ๋ฒ ํŒ…โ€™์„ ํ•˜์ง€ ์•Š๋„๋ก ์ฃผ์˜ํ•˜๊ณ  ํŒŒํŠธ๋„ˆ์‹ญ ๋‚ด์—์„œ ์ ์ ˆํ•œ ๊ทœ๋ชจ์™€ ์ฑ…์ž„์˜ ์ค‘์š”์„ฑ์„ ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค.

๋ธŒ๋ž˜๋“œ ๊ฑฐ์ŠคํŠธ๋„ˆ๋Š” AI ์Šคํƒ€ํŠธ์—…์ด ์–ด๋–ป๊ฒŒ ๊ณผ์—ด๋˜๊ณ  ์žˆ๋Š”์ง€, ๊ทธ๋ฆฌ๊ณ  ์†Œ์…œ ๋„คํŠธ์›Œํ‚น ์‹œ์žฅ๊ณผ ์œ ์‚ฌํ•˜๊ฒŒ ์—ญ์‚ฌ๊ฐ€ ๋ฐ˜๋ณต๋˜๋Š” ๊ฒƒ์„ ์–ด๋–ป๊ฒŒ ๋ฐ”๋ผ๋ณด๋Š”์ง€์— ๋Œ€ํ•ด ์ด์•ผ๊ธฐํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” AI์— ๋Œ€ํ•œ ๊ฑฐ๋Œ€ํ•œ ๋ฐ˜ ํฌํŠธํด๋ฆฌ์˜ค๋ฅผ ์–ธ๊ธ‰ํ•˜๋ฉฐ 60๊ฐœ ์ด์ƒ์˜ ํšŒ์‚ฌ์— ๋Œ€ํ•ด ๊ฑฐ์ ˆ ์˜์‚ฌ๋ฅผ ๋ฐํ˜”์ง€๋งŒ, ๋งŽ์€ ๊ฒฝ์Ÿ์—…์ฒด๊ฐ€ ์ด๋Ÿฌํ•œ ๊ฑฐ๋ž˜๋ฅผ ํ•˜๊ณ  ์žˆ์œผ๋ฉฐ ์ด๋Š” ์•„๋งˆ๋„ ํ‹ฐ์ € ๋ฒ ํŒ…์— ๋ถˆ๊ณผํ•  ๊ฒƒ์ด๋ผ๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋„คํŠธ์›Œํฌ, ๊ด€๊ณ„ ๋˜๋Š” ๋กœ๊ณ ๋ฅผ ๊ตฌ๋งคํ•˜๊ธฐ ์œ„ํ•ด ์ž‘์€ ๋ฒ ํŒ…์„ ๋งŽ์ด ํ•˜๋Š” ๊ฒƒ์€ ์†Œ์…œ ๋„คํŠธ์›Œํ‚น์—์„œ ํšจ๊ณผ๊ฐ€ ์žˆ์—ˆ๋˜ ๊ฒƒ์ฒ˜๋Ÿผ ์ด๋ฒˆ์—๋„ ํšจ๊ณผ๊ฐ€ ์—†์„ ๊ฒƒ์ด๋ผ๊ณ  ๊ฒฝ๊ณ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋ฒค์ฒ˜ ํŽ€๋“œ์˜ LP๊ฐ€ ๋ˆ์„ ์žƒ๋Š” ๊ฒƒ์— ๋Œ€ํ•ด ์šฐ๋ คํ•˜๋ฉฐ GP๊ฐ€ ์–ด๋–ป๊ฒŒ ์ด๋Ÿฐ ์ผ์ด ์ผ์–ด๋‚˜๋„๋ก ๋‚ด๋ฒ„๋ ค๋‘๊ณ  ์žˆ๋Š”์ง€ ๋ฌป์Šต๋‹ˆ๋‹ค.

Altimeter Capital Management์˜ CEO์ธ ๋ธŒ๋ž˜๋“œ ๊ฑฐ์ŠคํŠธ๋„ˆ๋Š” AI ์Šคํƒ€ํŠธ์—…์— ๋Œ€ํ•œ ๊ณผ๋„ํ•œ ํˆฌ์ž์— ๋Œ€ํ•œ ์ž์‹ ์˜ ์ƒ๊ฐ์„ ๊ณต์œ ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ์Šคํƒ€ํŠธ์—…์˜ ์ปดํ“จํŒ… ๊ตฌ๋งค์— 1์–ต ๋‹ฌ๋Ÿฌ์™€ ๊ฐ™์€ ๊ฑฐ์•ก์„ ํˆฌ์žํ•˜๋Š” ๊ฒƒ์€ ํ˜์‹ ์ ์ธ IP๋ฅผ ๊ตฌ๋งคํ•˜๋Š” ๊ฒƒ์ด ์•„๋‹ˆ๋ผ ์„ค๋น„ ํˆฌ์ž์— ๋ณด์กฐ๊ธˆ์„ ์ง€๊ธ‰ํ•˜๋Š” ๊ฒƒ์œผ๋กœ ์ˆ˜์ต๋ฅ ์ด ๋‚ฎ์„ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์กฐ์–ธํ•ฉ๋‹ˆ๋‹ค. ๊ฒŒ๋ฅด์ŠคํŠธ๋„ˆ๋Š” ํš๊ธฐ์ ์ธ IP๋ฅผ ๊ฐœ๋ฐœํ•˜๋Š” ์Šคํƒ€ํŠธ์—…์— ํˆฌ์žํ•ด์•ผ ์‹ค์งˆ์ ์ธ ์†Œ์œ ๊ถŒ์„ ํ™•๋ณดํ•˜๊ณ  ์ฐจ๋ณ„ํ™”ํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ์„ค๋ช…ํ•ฉ๋‹ˆ๋‹ค. ๊ทธ๋Š” ๋‹ฌ๋Ÿฌ๋‹น 80์„ผํŠธ๊ฐ€ ์˜์—…๊ณผ ๋งˆ์ผ€ํŒ…์— ์‚ฌ์šฉ๋˜๋ฉด ์ˆ˜์ต๋ฅ ์ด ๋‚ฎ์•„์ง„๋‹ค๊ณ  ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ๊ฑฐ์ŠคํŠธ๋„ˆ๋Š” ์Šคํƒ€ํŠธ์—…์— 1์–ต ๋‹ฌ๋Ÿฌ๋ฅผ ํˆฌ์žํ•˜๋Š” ๊ฒƒ์€ ๊ธฐ๋Šฅ์„ ์ž„๋Œ€ํ•˜๋Š” ๊ฒƒ๊ณผ ๋น„์Šทํ•˜๋‹ค๋Š” ์ ์„ ์ดํ•ดํ•ด์•ผ ํ•˜๋ฉฐ, ์Šคํƒ€ํŠธ์—…์€ ํ•„์š”ํ•œ ์ž๊ธˆ์„ ์กฐ๋‹ฌํ•˜๊ธฐ ์œ„ํ•ด Comerica Bank์™€ ๊ฐ™์€ ๊ธˆ์œต ๊ธฐ๊ด€์„ ์ฐพ์•„์•ผ ํ•œ๋‹ค๊ณ  ๋ง๋ถ™์ž…๋‹ˆ๋‹ค.

๋ธŒ๋ž˜๋“œ ๊ฑฐ์ŠคํŠธ๋„ˆ์™€ ์ฐจ๋งˆ์Šค ํŒ”๋ฆฌํ•˜ํ”ผํ‹ฐ์•ผ๊ฐ€ AI ์Šคํƒ€ํŠธ์—… ์—…๊ณ„์˜ ์ž๊ธˆ ์กฐ๋‹ฌ ์—ดํ’๊ณผ ์ด๋กœ ์ธํ•ด ๊ธˆ์œต ๋ฌธ๋งน์ด ์–ด๋–ป๊ฒŒ ์ƒ๊ฒจ๋‚˜๊ณ  ์žˆ๋Š”์ง€์— ๋Œ€ํ•ด ์ด์•ผ๊ธฐํ•ฉ๋‹ˆ๋‹ค. ๊ฒŒ๋ฅด์ŠคํŠธ๋„ˆ๋Š” ์ข‹์€ ๋น„์ฆˆ๋‹ˆ์Šค์— ๋งŽ์€ ์ž๊ธˆ์ด ์ง€์›๋˜๊ณ  ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•˜์ง€๋งŒ, ํŒ”๋ฆฌํ•˜ํ”ผํ‹ฐ์•ผ๋Š” ์ˆ˜์‹ญ์–ต ๋‹ฌ๋Ÿฌ๋ฅผ ์Ÿ์•„๋ถ“๋Š” ๊ฒƒ์€ ์ œํ’ˆ ๋ฐ ์œ ํ†ต ๋ฌธ์ œ๋ฅผ ์•ผ๊ธฐํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.

AI ์Šคํƒ€ํŠธ์—…์— ๋Œ€ํ•œ ํˆฌ์ž์˜ ์–ด๋ ค์›€๊ณผ ๋ชจ๋ธ ๊ฐœ๋ฐœ ๋ฐ ํ•™์Šต์ด ์–ผ๋งˆ๋‚˜ ๋น ๋ฅด๊ฒŒ ๊ฐ€์†ํ™”๋˜๊ณ  ์žˆ๋Š”์ง€, ๋น„์šฉ ๊ณก์„ ์ด ์–ด๋–ป๊ฒŒ ์›€์ง์ด๋Š”์ง€ ์ดํ•ดํ•˜๋Š” ๊ฒƒ์ด ์ค‘์š”ํ•˜๋‹ค๋Š” ์ ์„ ๊ฐ•์กฐํ•ฉ๋‹ˆ๋‹ค. ๋จธ์‹ ๋Ÿฌ๋‹์˜ ๋น„์šฉ ์ ˆ๊ฐ ์†๋„๊ฐ€ ๋ฌด์–ด์˜ ๋ฒ•์น™๋ณด๋‹ค ํ›จ์”ฌ ๋น ๋ฅด๊ธฐ ๋•Œ๋ฌธ์— AI ์Šคํƒ€ํŠธ์—…์˜ ๋น„์ฆˆ๋‹ˆ์Šค ๋ชจ๋ธ์€ ์‹œ์žฅ์— ๊ฐ€์žฅ ๋จผ์ € ์ถœ์‹œํ•˜๊ณ  ์ƒˆ๋กœ์šด ๋ฐ์ดํ„ฐ๋ฅผ ์ƒ์„ฑํ•˜๋ฉฐ ์ง€์†์ ์ธ ์ด์ ์„ ๊ตฌ์ถ•ํ•˜๋Š” ๋ฐ ์œ ๋ฆฌํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๊ธฐ์ˆ  ์šฐ์œ„ ์‹œ์ ์ด ์–ธ์ œ๊ฐ€ ๋ ์ง€ ๋ถˆํˆฌ๋ช…ํ•˜๊ณ , ํˆฌ์ž… ๋น„์šฉ์˜ ์ตœ์†Œํ™”๊ฐ€ ์กฐ์ • ๋น„์šฉ์—๋„ ์˜ํ–ฅ์„ ๋ฏธ์น˜๊ธฐ ๋•Œ๋ฌธ์— ๊ธฐ์ˆ  ํˆฌ์ž์— ์–ด๋ ค์›€์„ ๊ฒช๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ, ์Šคํƒ ์•„๋ž˜์—์„œ ๊ฒฝ์Ÿํ•˜๋Š” ๊ฒƒ๋ณด๋‹ค ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜๊ณผ ๋„๊ตฌ ๊ณ„์ธต์—์„œ ํ˜์‹ ํ•˜๊ณ  ํŒŒ๊ดดํ•˜๋Š” ๊ฒƒ์ด ๋” ์‰ฌ์šฐ๋ฉฐ, ์ด๋Š” ๋” ๋งŽ์€ ์ž๋ณธ์ด ํ•„์š”ํ•˜๊ณ  ์ˆ˜์ต์ด 0์ด ๋  ์ˆ˜๋„ ์žˆ๋‹ค๊ณ  ๋งํ–ˆ์Šต๋‹ˆ๋‹ค.
https://twitter.com/chamath/status/1670686049863979009

What is the biggest business secret hiding in plain sight right now?

Well, weโ€™ve been saying for a while on
@theallinpod
that we should prepare for interest rates โ€œhigher than we like and for longer than we wantโ€.

The most under reported issue in business is how Corporate America is about to hit a massive debt wall because of it.

A โ€œdebt wallโ€ is a term that describes how much debt is due at various times. If a lot of debt comes due at the same time, it puts a lot of pressure on a company.

It turns out that companies issued a ton of short term debt during the pandemic at close to 0% interest rates. As you can see below, hundreds of billions of dollars will come due starting Jan-2024 and will need to be refinanced at MUCH higher rates.

Prepare for a bunch of companies who will not be able to refinance their debt and will thus see their equity value incinerated. This will hit the private equity industry very acutely, whose core playbook involves wrapping their companies in gobs of high yield debt.

That said, this will also ultimately create opportunities for those with capital on the sidelines to act as a buyer and recap the best of these companies.

Expect a bunch of articles about potential corporate bankruptcies starting this fallโ€ฆlots of money to be made if youโ€™re paying attention.