팔란티어의 온톨로지: AI 수익 및 의사 결정의 혁신

### 팔란티어의 온톨로지: AI 지배력과 매출 성장의 열쇠: 팔란티어의 온톨로지
* 🔍 팔란티어는 컴퓨터 처리를 위해 데이터 관계를 구성하고 분석하는 온톨로지에 중점을 둔 **Apollo 및 Foundry와 같은 고가의 플랫폼**에 특화되어 있습니다.
* 💡 온톨로지는 데이터 품질을 보장하고 AI 개발의 중요한 과제인 잘못된 정보를 방지하기 때문에 **AI 정확도에 매우 중요**합니다.
* 🚗 예를 들어, 자율주행에서는 양질의 데이터가 필수적인데, **엘론 머스크는 정지 신호에 완전히 정차하는 사람은 5%에 불과하다**며 선별된 데이터의 필요성을 강조했습니다.
* 🏢 팔란티어의 온톨로지 전문성은 **현재 진행 중인 러시아와의 분쟁에서 우크라이나 군을 지원**하는 등 다양한 분야에서 활용되고 있습니다.
* 🔄 기존에는 파견 개발자를 통해 온톨로지를 설정하는 데 **6개월**이 걸렸지만, **팔란티어의 새로운 AI 플랫폼(AIP)은 이를 단 며칠**로 단축했습니다.
* 💼 팔란티어는 **보안 승인(영향 수준 6)과 정부 계약**을 통해 기밀 데이터를 처리하는 데 있어 경쟁 우위를 확보하고 있습니다.
* 📊 팔란티어는 온톨로지 및 AI 구현을 선도하고 있지만, **빅테크(예: Microsoft, Amazon, Meta)와의 경쟁**은 잠재적인 위험을 초래할 수 있습니다.
* 📈 팔란티어의 향후 실적에 대한 주식 시장의 엇갈린 의견은 **빠르게 진화하는 AI 산업에서 경쟁 우위에 대한 불확실성**을 반영합니다.

팔란티어는 효과적인 의사 결정을 위해 온톨로지를 통해 데이터를 조직화하여 세계 최고의 AI 수익 기업이 되는 것을 목표로 합니다.

If you invest in American stocks, you have probably heard of Palantir at least once. It holds the title of being the most secretive company in the world, so many investors may not know what Palantir really does and invest without knowing. People think it looks cool even if they roughly search or watch videos. I also invested after a rough look and ended up cutting losses. However, today, I will take a very interesting look at why Palantir is receiving so much attention for its AI and how it can become the world's number one in AI field revenue. Hello, I'm the Tech Farmer. First of all, Palantir is a company that sells high-priced platforms like Apollo and Foundry that we are good at. Understanding each platform is good, but we understand the word that is the basis of all these platforms. That word is ontology.

In simple terms, it is organizing and analyzing the relationships between data and representing them in a form that computers can process. If you listen to the end of this video, I believe you will understand what it means. Let's take a car as an example. If there is an object called a car, we can express how fuel-efficient this car is and how many people can ride in it. Also, if there is an object called driving, we can define the time of departure and arrival, and if there is an object called a parking lot, it has characteristics such as when it was built and how many cars it can accommodate. In other words, organizing objects related to each other, such as cars, driving, and parking lots, into a hierarchical rule is called an ontology. It's about giving data limits and categorizing them. It may not be easy to understand right away, but if you keep watching, you will be able to understand. And ontology...

The collected data contains information that is not abstract but can be agreed upon by anyone. For example, the fuel efficiency of a car and the number of passengers are data that anyone would consider correct. I tried to use ChatGPT for YouTube video production in the past because it took more time to determine whether the ChatGPT responses were real or fake. This is because ChatGPT is not yet fully ontologized. The ontology seen earlier contains information that anyone would consider correct. There is not a single incorrect feature among its characteristics. Therefore, when asking a question, even if the correct answer is given, there may be subtly incorrect information inserted in between. For example, when I asked to be informed about who General Yi Sun-sin is, it is written at the end that even his tomb has not been found. When we see this, we still do not know who General Yi Sun-sin is.

It may seem that the tomb cannot be found, but in Asan City, Chungcheongnam-do, there is actually the tomb of General Yi Sun-sin. This means that there is some fake information mixed into the data of General Yi Sun-sin. However, currently, AI cannot distinguish between the real and fake parts. That is why ontology is important because it needs to define concepts clearly as actual existing words. Looking at ontology and the given data, if all 100 out of 100 people say it is right or wrong, it can be considered data. When abstract or virtual concepts are extracted and input into the computer, it is necessary to verify whether the data is correct before inputting. For example, when considering the statement Will your head become smarter through this training method?, will everyone say the same thing about how to make your head smarter? Also, when interpreting the meaning of being smart when young as a student and being smart after aging, are these two interpretations the same?

In the past, being good at studying was considered smart, but nowadays, people who speak well at work or earn money through investments can be seen as smart. For example, some may argue that not smoking can increase concentration and make a person smarter, but this claim has not been proven in any research papers. This is how fake information can spread. Without clear rules and ontology, we are more likely to fall into misinformation and speak incorrectly. It's like what we saw earlier, like a teacher. Palantir Ontology aims to correct these issues and build a digital twin. A digital twin is a representation of the real world transformed into digital form and implemented in the computer space. In other words, all data should be inside the computer. If there is no ontology, a digital twin cannot be built.

In the online world, fake information and real information will be mixed, so Palantir emphasizes the importance of ontology and is building a digital twin. This way, it can help deal with fake information, accurately understand the company's issues, and assist in making the right decisions. Let's take a look at an actual event at Tesla, for example. Recently, Elon Musk showed a scene of autonomous driving with the Beta 12 version live. Starting from this Beta version, autonomous driving is not done through code but by training AI with video data. In other words, it's not that the command go on a green light is inputted through code, but rather, the AI learns from numerous video data where people start on a green light and makes decisions on its own. From this live video, we can see points related to Palantir. This means that even with a vast amount of ordinary data,

It's saying that good quality data is needed because it doesn't help in the driving room, meaning that quality data needs to be filtered out. So, there is separate software for filtering quality data. In this video, Elon Musk is mentioning an interesting example. One of the slightly funny challenges we've had is that, since the car is being trained on what humans do, humans almost never fully stop at a stop sign. So when they get to a stop sign, humans actually almost never go to zero miles an hour. They may think they did, but usually, they are doing at least a few miles an hour coming up to a stop. Sometimes that bothers the regulators. Some quite insist that we go to a complete stop at stop signs. When we looked at the data, only 5% of the time do humans actually stop, even lower than that, 5%. Wow. So basically, people almost never fully stop at stop signs. In Korea, cars should come to a complete stop at bus stops, but generally, Americans do not stop at the stop sign. In Korea, when making a right turn, you should stop in front of the pedestrian crosswalk and then proceed, but most people just go. In other words, if you input AI to just enable autonomous driving.

Autonomous driving can stop at bus stop signs and avoid accidents. As mentioned earlier, autonomous driving must be done safely, but if it is not operated safely, fake data can be generated, so it is necessary to collect high-quality data that can be used at bus stops and train AI. They talk about curation. Data curation means selecting useful data. Ontology may be confusing. As mentioned earlier, to curate data, which is the process of acquiring data, you need to go through the ontology process that enriches the meaning of the data. In other words, even Elon Musk mentioned that ontology is difficult. No matter how excellent Tesla's AI technology is in the autonomous driving market, if a lot of cost and time are spent on ontology like this, the development of robotaxis and Tesla's vision may slow down. And Elon.

What if Musk is managing Palantir by forming a partnership to distinguish high-quality data? There are rumors that Tesla and Palantir have actually formed a partnership. Please subscribe and like. If you look now, you can see a domain with Tesla's name added to Palantir's domain. This is a domain created by Palantir. Just for your information, Palantir doesn't have a free version or a test version; you must sign a contract to provide the domain and Foundry software. Alternatively, it could simply be a test site created by Palantir employees. Nonetheless, when Elon Musk and Palantir CEO Peter Thiel heard about the PayPal Mafia, they could see how challenging the software developed by Palantir is. They could understand how difficult the software developed by Palantir is, considering the software Palantir is working on. The world's top autonomous driving company's AI supercomputer, Dojo.

Despite the difficulties, Palantir software can be seen to help customers select and give meaning to numerous data tailored to their goals, enabling effective decision-making by deriving conclusions. Another example is the ongoing war in Ukraine. How has Ukraine been able to continue the war against Russia, one of the world's most powerful military forces? It is because of Palantir. By comprehensively analyzing numerous satellite data, information on the Russian military, and the dynamics of the war so far, Palantir has been able to inform the Ukrainian military of possible locations of the Russian army. By attributing meaning and categorizing the numerous sporadic data that arise during the war, ultimately, it extracts the desired goal of the customer. In this case, the goal is precisely the location of the Russian army. It is similar to using a map pack in the game StarCraft.

Using this kind of technology to generate revenue, Palantir can be considered almost the only company doing so. However, Palantir had a problem with Ontology. It was difficult for customers to build it directly. Because customers need to input the exact purpose they want into the Palantir program, someone who knows the company or industry well needs to set up the Ontology. However, customers know what they want, but they have no idea how to Ontologize it in the Palantir program. Therefore, Palantir sends FDE developers to the customer companies until the customers can build the Palantir program themselves. The picture now is with an FDE dispatched to Doosan Infracore. The person in blue is the Palantir FDE. While staying at the company, they understand the flow of the industry.

At the same time, we explain how to build a customized ontology for the company and how to utilize software. Therefore, developers spend a long time at the company. This means that they have to customize each one for Palantir customers. Therefore, the biggest problem with this method is that it takes too long, making mass production difficult. Customers who initially signed contracts with Palantir cannot use Palantir programs immediately. It takes time to build them through dispatched developers. Even after the dispatched developers have built it, they continue to manage it due to maintenance issues. It usually took about six months. However, last year, Palantir made this announcement: In response to the advancement of AI, Palantir introduced the Palantir AI platform AIP, which supports customers to use large-scale language models along with existing machine learning technologies.

It was released. During the performance announcement in the third quarter of last year, the Palantir CEO said that by using AIP, the ontology construction that used to take six months with existing dispatched developers was completed in just a few days. Previously, because ontologies differed by industry, FDE had to be sent for customization one by one. However, by using AI to do the customization work, it is possible to complete the customization in a short period. In this way, Palantir can sell the program in large quantities. In fact, after Palantir introduced AIP in June last year, on August 28th, an analyst from Bank of America acknowledged Palantir's capabilities through various customers and analyzed that they are in a unique position to enable AI implementation. However, the current target price of institutions mostly shows a sell recommendation at $1. For Palantir, solving companies' problems in the ontology.

The reason why Palantir was able to quickly enter the market that solves problems is due to its complementarity. To do ontology, you need to analyze all the data of companies, which of course includes confidential information. However, Palantir gained recognition by handling such confidential information in a way that customers can trust. There are only three companies that have been granted Impact Level 6, the highest level of handling classified data by the U.S. Department of Defense, and Palantir, Microsoft, and Amazon are among them. The fact that they started by targeting U.S. government agencies, not private companies, shows how excellent their stability and complementarity are. However, is Palantir's software truly unique and exceptional? No matter how much they have been in the field for 20 years, the current market is growing rapidly in AI, so we can see how excellent Palantir's stability and complementarity are. Nevertheless, Palantir is dominant in the software industry and truly unique.

I also think that following Big Tech could pose a significant risk to Palantir's stock price. Along with Palantir, let's take a look at Microsoft and Amazon, both of which can handle the highest-level classified information from the U.S. Department of Defense. Microsoft recently provided a feature in its GPT-integrated CoPilot that enables quick decision-making. Furthermore, using CoPilot's AI library, it is said to promote more strategic, data-driven thinking and decision-making. At the 2022 re:Invent event, the Vice President of AWS Data and Machine Learning announced that they have added features that allow customers to make data-driven decisions easily. In other words, Big Tech companies have started to incorporate features that can help solve customer problems, as mentioned earlier in the case of Palantir Ontology. However, as seen before, even Tesla is curating data to solve problems but still faces many issues. Amazon and SpaceX also have similar challenges.

I think there will be problems. First of all, it is because it is impossible to distinguish good data as well as Palantir. However, recently, Meta is trying to advance in the AI market with a huge amount of money, such as purchasing 350,000 NVIDIA chips. Therefore, if Palantir creates a program that automatically ontologizes like Palantir AI feature, it could have a significant impact on Palantir's revenue. Currently, this is considered a dominant task in this field, but because big companies can easily catch up with money, it is essential to consider performance in investing in Palantir. The information I have explained so far may be somewhat subjective or abstract when it comes to investing in stocks, but performance data can be objectively viewed because it is represented by precise numbers. Therefore, in the next video, let's dissect all of Palantir's performance and analyze its future growth.

I have analyzed a company after a long time. I will upload the second video by analyzing it as quickly as possible. Have a great day. Thank you. Please note that this is not a recommended video, but my personal opinion. I have opened a Naver Premium Content channel where I share my personal stories, stock trading records, Korean stocks, real estate, and the process of actually making money that I couldn't cover on YouTube. Thank you for watching my videos so far. Please subscribe and like.

일론머스크도 팔란티어의 기술이 필요하다.

안녕하세요. 재테크 농부입니다.

24년 3월 네이버 프리미엄 콘텐츠 주간 인기 채널 전체 1위 달성.
24년 3월 네이버 프리미엄 콘테츠 미국 주식 기회전 10대 채널 선정.
24년 3월 ~ 5월 네이버 프리미엄 콘텐츠 재테크 분야 1위 달성.

https://contents.premium.naver.com/finfarmer00/finfarmer

저는 주식 전문가가 아니기 때문에 여러 자료를 모아 제 생각과 함께 분석한 영상입니다.

혹시나 잘못된 점이 있으면 댓글로 언제든 말씀 부탁드립니다.

또한 궁금한 기업이 있으면 댓글 남겨주시면 제가 공부해서 공유해 드리겠습니다.

개인적인 의견일 뿐 매수 추천은 절대 아닙니다.

투자에 대한 책임은 본인에게 있습니다.

비방, 정치, 분란, 예의 없는 댓글들은 경고 없이 삭제 및 차단합니다.

문의 : finfarmer00@gmail.com

좋은 하루 되세요!!



재테크 농부

팔란티어의 비밀스러운 세계와 AI 수익 잠재력

세계에서 가장 비밀스러운 회사로 알려진 팔란티어는 AI 역량과 AI 분야 수익의 선두주자가 될 수 있는 잠재력으로 주목받고 있습니다.

이 회사는 데이터 관계를 구성하고 분석하는 온톨로지에 중점을 둔 Apollo 및 Foundry와 같은 고가의 플랫폼을 판매합니다.

온톨로지를 사용하면 데이터를 분류하고 구조화하여 보편적으로 합의된 정보를 제공할 수 있습니다.

이러한 잠재력에도 불구하고 ChatGPT에서 볼 수 있듯이 팔란티어의 AI는 완전히 온톨로지화되지 않았기 때문에 때때로 미묘하게 잘못된 정보를 제공하기도 합니다.

정보 관리에서 온톨로지의 중요성

이순신 장군의 묘는 실제로 충남 아산시에 위치해 있지만, 그의 묘소에 대한 가짜 정보가 존재합니다.

온톨로지는 특히 디지털 공간에서 개념을 명확하게 정의하고 진짜와 가짜 데이터를 구분하는 데 매우 중요합니다.

팔란티어는 가짜 정보를 효과적으로 관리하고 기업의 문제를 파악하며 의사 결정을 지원하기 위해 디지털 트윈을 구축하는 데 온톨로지의 중요성을 강조하고 있습니다.

자율 주행에서 고품질 데이터의 중요성

엘론 머스크가 강조한 것처럼, 고품질 데이터는 자율 주행에서 AI를 학습시키는 데 매우 중요합니다.

인간은 정지 신호에 완전히 멈추는 경우가 드물기 때문에 AI가 신뢰할 수 있는 데이터를 통해 학습해야 할 필요성이 강조됩니다.

데이터 큐레이션과 온톨로지 강화 과정은 어렵지만 유용한 데이터를 확보하는 데 필수적입니다.

테슬라와 팔란티르가 고품질 데이터를 확보하기 위해 파트너십을 맺었다는 소문이 돌면서 자율주행 기술 개발에 영향을 미칠 가능성이 있습니다.

데이터 분석 및 의사 결정에서 팔란티어의 역할

팔란티어 소프트웨어는 데이터를 분석하고 의미를 부여하여 고객이 효과적인 의사결정을 내릴 수 있도록 지원합니다.

우크라이나-러시아 전쟁에서 팔란티어는 우크라이나 군에 러시아군의 가능한 위치에 대한 정보를 제공하는 데 결정적인 역할을 했습니다.

이 회사는 고객을 위한 온톨로지를 구축하는 데 어려움을 겪었지만 대규모 언어 모델과 머신 러닝 사용을 지원하기 위해 팔란티르 AI 플랫폼을 도입했습니다.

팔란티어의 AIP 영향력 및 시장 지위

팔란티어의 AIP는 온톨로지 구축에 혁신을 일으켜 사용자 정의 시간을 획기적으로 단축하고 대량 프로그램 판매를 가능하게 했습니다.

기밀 데이터를 처리하는 이 회사의 독보적인 능력은 특히 미국 정부 기관으로부터 인정과 신뢰를 얻었습니다.

빅 테크와의 경쟁에 직면해 있지만, 팔란티어는 안정성과 상호보완성을 바탕으로 빠르게 성장하는 AI 시장에서 차별화되고 있습니다.

메타의 AI 발전이 팔란티어의 수익에 미치는 잠재적 영향

메타의 대규모 NVIDIA 칩 구매를 포함한 AI에 대한 막대한 투자는 팔란티어의 수익에 직접적인 영향을 미칠 수 있습니다.

AI 시장의 경쟁이 치열해지면서 대기업들이 자금력을 앞세워 팔란티어를 따라잡고 있기 때문에 투자에 앞서 팔란티어의 성과를 고려하는 것이 중요합니다.

객관적인 성능 데이터는 다음 비디오 분석의 초점이 될 것이며, 이를 통해 팔란티어의 미래 성장을 평가할 수 있는 보다 구체적인 근거를 제공할 것입니다.