다가오는 물결: AI와 합성 생물학의 도전 과제

무스타파 술레이만은 AI와 합성 생물학이 인류를 변화시켜 전례 없는 기회와 심각한 위험을 동시에 가져올 수 있다고 경고합니다.

### 다가오는 물결: AI와 합성 생물학의 미래 탐색하기
* 🤖 알파고를 만든 것으로 유명한 딥마인드는 데미스 카사비스와 무스타파 술레이만이 공동 설립한 회사로, 후자는 **AI와 합성 생물학의 중요한 발전을 다룬 책 '다가오는 물결'**을 저술하기도 했습니다.
* 📚 저자는 현대 문명의 근간을 이루는 **인공지능, 생명공학, 에너지 등 세 가지 분야의 중요성**을 강조합니다.
* 🌊 이 책은 **AI와 합성 생물학이 새로운 부와 풍요의 시대**를 열어줄 것이지만, 제대로 관리하지 않으면 **혼란과 재앙의 위험**도 있다는 점을 강조합니다.
* 🔍 GPT-4와 같은 AI 기술은 **코드 생성, 프레젠테이션 초안 작성, 법률 문제 지원**과 같은 복잡한 작업까지 수행할 수 있어 다양한 산업 분야에서 혁신적 잠재력을 보여주고 있습니다.
* 합성 생물학은 **제약 및 유전자 편집에 미생물을 사용하는 등 실제 사례**를 통해 글로벌 난제를 해결할 수 있는 방대한 잠재력을 보여줍니다.
* ⚠ 그러나 저자는 논란이 되고 있는 허젠쿠이 사건에서 보듯이 **생물학적 무기를 만들 수 있는 가능성**과 유전자 편집의 윤리적 함의 등 이러한 기술과 관련된 위험에 대해 경고합니다.
* 🌍 이 책은 AI와 합성생물학의 급속한 발전을 관리하기 위한 국제적 협력과 규제의 필요성을 논의하며, **경쟁이 치열한 글로벌 환경에서 일방적인 조치만으로는 충분하지 않을 수 있다**고 강조합니다.
* 📖 궁극적으로 '다가오는 물결'은 **AI와 합성 생물학이 가져올 심오한 변화**를 이해하고 대처하는 동시에 관련 위험과 윤리적 책임을 고려하기 위한 행동 촉구 역할을 합니다.

There is a company called DeepMind, which is known for creating AlphaGo, the program that defeated Go champion Lee Sedol. The person on the left in the photo is Demis Hassabis, the founder of DeepMind, and he is the most well-known figure associated with the company. On the right is Mustafa Suleyman, who is also a co-founder of the company but is relatively less known. The book we will look at today, The Coming Wave, is written by Mustafa Suleyman. Since he works at the forefront of AI, his warnings about AI do not fall on deaf ears. Now, let's get into the content. Recently, over the past year or two, I have been examining many books related to technology and those related to the brain. If I were to pick the three fields I consider most important, the first would be artificial intelligence, and the second would be biotechnology, or...

The third field related to life sciences is the energy sector, including technologies that address climate change. Ultimately, what constitutes modern civilization are life, intelligence, and energy. It might be helpful to keep these three concepts in mind. Furthermore, this book focuses particularly on intelligence and life within the framework of life, intelligence, and energy. Let's listen to the author's words. The upcoming wave is defined by two key technologies: artificial intelligence (AI) and synthetic biology. These two technologies will open a new dawn for humanity and create unprecedented wealth and abundance. However, the rapid spread of these technologies could also empower individuals who engage in miraculous acts to unleash chaos, instability, and even disaster on an unimaginable scale. This wave of change presents a tremendous challenge that will define the 21st century, and our future is...

While relying on these new technologies, we are also facing threats. Synthetic biology may sound a bit complicated, but it involves manipulating genes and redesigning living organisms. So, I affirm that the opportunities of the future will lie in these fields. Yes, and crises will also arise from these areas. 

First, let's look at artificial intelligence. Today’s AI can write long texts, generate images and sounds, and recognize human faces and objects. Although it is not perfect yet, it can also drive autonomously. There are direct applications of AI, but this technology drives advancements in other technological domains as well. For example, advancements in genetic technology, robotics, quantum computing, and energy technology are all intertwined with the development of AI, leading to tremendous progress.

To illustrate, let's talk about computer vision technology related to object recognition. In 2016, Amazon introduced a cashier-less supermarket called Amazon Go. Customers who are signed up with Amazon can simply enter this supermarket, pick up items, and leave, with the payment being automatically processed and charged. Of course, we could discuss separately whether this has been successful in terms of profitability, but let's focus on the technology itself. The technology used here is computer vision, which trains artificial intelligence to recognize and interpret objects. This allows the system to identify actions such as a person picking up milk and placing it in a bag or grabbing a sandwich and putting it in a bag. It enables automatic calculation of the total cost for charging. Such technology is also applicable to drones.

It can also be used in technology to recognize and attack enemies. Additionally, in car manufacturing, robots could be equipped with vision systems. For example, when a robot is moving a steel car part and accidentally a human worker gets in the way, the robot could automatically recognize the person and adjust its movements. This technology could certainly be applied to autonomous vehicles as well. Although it's a grim thought, it could also be used in CCTV systems to monitor citizens. The key point here is that artificial intelligence can mimic various human functions, and I'm discussing how computer vision technology, which mimics human sight, can transform many industries. Next, I will talk about the performance of the language processing model, GPT-4. There have been articles stating that GPT-4 has become less intelligent, but I will do my best to address the author's argument.

I will follow along. AI GPT-4 can work with images and code, create 3D computer games that run on desktop browsers, develop smartphone apps, debug code, identify loopholes in contracts, suggest modifications to compounds to avoid patents, and even create websites from hand-drawn images. It understands the subtle dynamics of humans in complex scenes and can propose recipes based on the food in a refrigerator when shown one. When drafting a presentation, it can refine and design it to a professional level. GPT-4 seems to understand spatial causal reasoning, medicine, and law. Just days after the release of GPT-4, people created tools that automatically handle lawsuits, assist with joint custody of children, and provide real-time fashion advice. A few weeks later, people began developing mobile apps or conducting market research.

I have created additional features that allow for complex tasks to be performed in detailed report writing. The performance of this AI and its various functions are already well known, and I have introduced them many times. Next, I would like to discuss synthetic biology. I was considering how to explain this, and I found that looking at practical examples from the synthetic biology core technology development and dissemination strategy released by the Ministry of Science and ICT last October would be helpful. This will give you a sense of what synthetic biology encompasses. 

For example, the first point discusses using microorganisms to create pharmaceutical materials, which means that gut microorganisms can produce drugs. The second point is about designing antibodies using artificial intelligence. The third point involves developing therapeutics using gene editing tools, which I have introduced before. Additionally, there are efforts to convert methane, which can be considered a major contributor to global warming, into high-value materials by utilizing carbon dioxide. Following this,

Fifth, we will decompose plastic using microorganisms. Sixth, we will replace pesticides and nitrogen fertilizers with microorganisms. Next, seventh, we will develop bio-materials that mimic the taste of meat. Then, eighth, we will enable the rapid production of useful bio-materials on-site. Ninth, we will enhance the efficiency of photosynthesis in plants through gene editing. This field is diverse, and each aspect has the potential to bring about significant changes. Everything is under the trend of synthetic biology. When I encounter these topics, I feel a sense of anticipation for the massive changes that are approaching. You may have heard of Moore's Law, which states that the number of transistors on a semiconductor chip doubles approximately every two years. In the field of life sciences, the counterpart to Moore's Law is Carlson.

Here, DNA sequencing analysis, that is, reading DNA, is reportedly decreasing in cost at a faster rate than Moore's Law. While the cost of DNA synthesis is slower compared to DNA analysis, it is also showing a rapid decrease in cost. Originally, this was represented by a curve, but I have simplified it into a straight line. 

Now, let's talk about proteins. Typical proteins are made up of about 250 to 300 amino acids from twenty different types, intricately linked together. These amino acids fold and twist to determine the three-dimensional structure of the protein. Since changes in protein structure can alter its functions and activities, understanding protein structure is of great importance. Predicting protein structure is indeed a crucial task, but it is also quite challenging.

It was a significant development. However, AlphaFold has created an artificial intelligence that can solve this problem. AlphaFold has learned the sequences of amino acids and the structures of proteins, allowing it to predict protein structures based on this knowledge. In fact, AlphaFold demonstrated overwhelming predictive power and won a protein structure prediction competition. We have entered an era where artificial intelligence is applied in biotechnology. The database of protein structures predicted by DeepMind's AlphaFold is publicly available. The image on the left was also downloaded from that database. If we can understand the structure of protein folding well, we will be able to understand the process of disease occurrence better. Therefore, it is expected that there will be significant advancements in drug development and more in the future. The key point is that a substantial part of the essence of life—if we can call it that—lies in information. Thus, AI plays a crucial role.

As we progress, many things can also advance in the field of life sciences. I will read this once. At the center of this new wave is the recognition of a biologically evolved system that represents, encodes, and stores DNA information. Until now, all intelligence has come from living beings. It is sometimes referred to as synthetic intelligence or artificial life, but its meaning remains the same. Both fields are completely fundamental and interconnected concepts; that is, they are about recreating and designing two core attributes of humanity. If we change our perspective, these two fields become one project. Additionally, in synthetic biology, the scale of basic operations is very small, in areas like DNA. Therefore, the advancement of nanotechnology is inevitable. Originally, when we said a singularity would occur, the foundational disciplines were molecular biology and nanotechnology.

Biotechnology, followed by information technology, and then cognitive science—these four fields are emerging. In fact, these elements are shaping a massive trend. The reason the book is titled The Coming Wave can be attributed to this. The technologies of artificial intelligence and synthetic biology that I have explained so far are general-purpose technologies. They can be applied to various fields, and their impact is widespread. In human history, there have not been many technologies that can be considered general-purpose, but once they emerge, they dramatically change the course of human history. The stone axe did that, as did fire. The emergence of language, agriculture, and writing also completely altered the flow of history. AI and synthetic biology are expected to play a similar role. However, when these technologies are invented, unforeseen negative consequences can arise, which is known as the backlash effect.

Gutenberg simply wanted to print the Bible to make money. However, the printing press he invented triggered the scientific revolution and the Reformation, becoming the greatest threat since the establishment of the Catholic Church. Just as those who developed internal combustion engines and jet engines had no intention of melting glaciers, refrigerator manufacturers likely did not intend to create holes in the ozone layer with chlorofluorocarbons. Now, artificial intelligence and synthetic biology fall into the category of general-purpose technologies, and they don't just cause problems; they create meta-problems. Let's talk about what disasters these incredible technologies could bring. As I mentioned earlier, when drone technology is combined with computer vision and armed with offensive weapons, it becomes a very complex weapon. Today, there were several reports about the U.S. military in Jordan.

There was an incident where a drone attack resulted in three deaths and thirty injuries. Drones are now emerging as a key component of asymmetric warfare. If we imagine hundreds or thousands of such weapons flying over urban areas, it would be a truly terrifying situation. Additionally, last year, the Houthi rebels in Yemen attacked commercial vessels passing through the southern Red Sea, targeting ships associated with Israel using drones and missiles. Fortunately, since these targets were large ships, the scale of casualties was not significant. However, if these drones were to attack urban areas and were equipped with chemical or biological weapons, the damage would be unimaginable. There are various discussions regarding whether the coronavirus emerged naturally or was artificially created in a laboratory. For now, let's leave the coronavirus as an unknown territory.

As I mentioned, will the technologies of synthetic biology, which can be utilized in various fields, only be used for good purposes? It is hard to say so. If the technology of synthetic biology continues to advance intricately, it may be possible to artificially create new viruses that can cause epidemics. Moreover, the issue is that, unlike nuclear weapons, where movements can be somewhat predicted and international monitoring can be actively conducted, DNA synthesizers are becoming popular enough for experiments to be conducted by individuals, which poses a risk similar to someone being able to create nuclear weapons in their own home. If new viral pathogens were to be transported by drones to carry out attacks, they would become powerful weapons. Additionally, in 2018, there was an announcement by the Chinese scientist He Jiankui regarding the birth of twin babies through gene editing.

The technology applied was not verified for safety, and it was unethical. Subsequently, He Jiankui received tremendous criticism globally and was sentenced to three years in prison by the Chinese government, where he is nearing the end of his sentence. It seems he will be released soon. In any case, this gene editing is not a one-time event; the descendants of those twins will continue to be affected by the edited genes. Therefore, if gene editing is used not only for therapeutic purposes but also to enhance certain functions, it could lead to polarization and raise various issues in that field. Moreover, individuals with the intent to create fake news can now utilize AI with advanced writing capabilities to generate plausible stories that contain 1% falsehoods in their texts. This manipulation is not limited to text; videos and audio can also be convincingly and artificially altered.

We have entered an era where deep learning technology allows us to create fake videos and images. As this technology advances, its influence becomes enormous, leading to various problems. Naturally, it is essential to identify these issues and make efforts to address them. However, there are fundamental forces that create an atmosphere discouraging such efforts, insisting that technology should not be hindered. Recently, in Europe, laws regulating artificial intelligence have been enacted, undergoing approval processes in various countries. Amid this, some claims have emerged, arguing that Europe is making a mistake. Authoritarian countries like China are advancing technology without regard for human rights or personal data privacy, while Europe seems to be tying its own hands. Such claims arise when considering the international order characterized by the collision and competition of these major powers.

There can be a strong flow that suggests this is not the answer. Economic incentives also exist. Emerging technologies like AI and synthetic biology will have a significant impact on almost every field. The ripple effects on national economic growth rates, corporate profits, and citizens' quality of life are enormous. There are incentives to follow technological competition, such as geopolitical competition and economic ripple effects, which create an atmosphere that undermines measures to suppress technology. In the fourth part of the book The Coming Wave, ten measures for suppressing technology are outlined. There are various approaches, but when looking at a single country, the role of state power and citizens is likely to be important. This topic can be referenced in the earlier work by Darren Esmond Glu, Power and Progress.

There will be a need for critics from the production department to participate, especially critics who have a practical focus. In my opinion, the most important thing ultimately seems to transcend national boundaries. While we can manage internal issues and their side effects in any way, it is much more challenging to handle when other countries are being greedy. Even if Europe or the United States has good regulatory measures and strives for balanced development, if, for instance, China is speeding ahead without brakes, Europe and the United States will simply fall behind. Therefore, I believe it is crucial to consider measures that take into account the international order. The author also presents this idea. The third key point is to impose export controls on advanced materials and technological components. This aims to delay the arrival of new waves in the region and buy time necessary for these regulations.

It is explained that in the case of the alliance of number seven, the possibility of cooperation among great powers such as the United States and China in synthetic biology exists. This raises concerns about the outcomes of synthetic biology that could potentially destroy opponents, as well as the undeniable threat posed by terrorist attacks. Therefore, there is an incentive to establish an international cooperation system in this field. I often look for books written by key figures who are active at the forefront of the tech industry about the direction in which this world and society are heading or should head. If those who benefit most from technological advancements feel a sense of urgency regarding these developments, I believe their insights are worth listening to. So today, we will discuss artificial intelligence.

I briefly looked into the book The Coming Wave by Mustafa Suleyman, who works on the front lines. It discusses the significant changes that AI synthetic biology will bring, as well as the potential disruptive impacts and the measures to mitigate them. I believe it covers a vast amount of content. Although the book is thick, the material is not particularly difficult, making it a quick read. I thought it would be necessary to revisit the ten measures for mitigation when we encounter this issue again in the future. So today, I introduced Mustafa Suleyman's book The Coming Wave. Let's wrap it up here. Thank you.

딥 러닝의 공동창립자 무스타파 술레이만의 책 『더 커밍 웨이브』

[BASE CAMP] "기술에 대한 철학적 조망은 우리의 사업과 투자, 더 나아가 우리의 삶이 가야 할 길을 인도해 줄 것이다."

우주 공간에서 아름다운 지구를 바라본 우주비행사들은 완전히 다른 인간이 되어 이후의 삶을 살아가게 됩니다. 이를 조망 효과라 합니다.  오쏘는 추상적 학문들의 힘을 빌려 지상에서의 조망 효과를 추구합니다.

Layer1: 철학(기술철학, 과학철학, 심리철학, 현상학), 과학(사이버네틱스, 인지과학, 신경과학)
Layer2: 기술(IT, BT, AI)
Layer3: 국제정세, 지정학, 투자, 사업, ESG, 자기 계발

문의: ooyyjj11@naver.com으로 부탁드립니다.


오쏘의 조망 효과 - "지식을 담다"

알파고로 유명한 딥마인드는 데미스 하사비스와 무스타파 술레이만이 설립한 회사로, 술레이만은 저서 '다가오는 파도'에서 인공지능의 잠재적 위험성에 대해 경고한 바 있습니다.

저자는 현대 문명을 형성하는 데 있어 인공지능, 생명공학, 에너지 기술의 중요성을 강조합니다.

AI와 합성 생물학의 급속한 확산은 전례 없는 기회와 잠재적 혼란을 동시에 가져와 유전자 기술, 로봇 공학 등 다양한 기술 영역의 발전을 이끌고 있습니다.

인공지능은 로봇 공학, 자율주행차, CCTV 시스템 등에 적용되어 인간의 기능을 모방함으로써 다양한 산업을 변화시키고 있습니다.

이미지 및 코드 처리, 3D 게임 개발, 앱 제작, 심지어 법률 및 의료 지원까지 가능합니다.

합성 생물학은 약물 생산을 위한 미생물 활용, AI를 이용한 항체 설계, 치료용 유전자 편집, 글로벌 난제 해결을 위한 바이오 소재 개발과 같은 이니셔티브를 통해 큰 변화를 가져올 것입니다.

AI와 합성 생물학이 생명 과학에 미치는 영향

DNA 시퀀싱 비용은 무어의 법칙을 앞지르며 빠르게 감소하고 있으며, DNA 합성 또한 빠른 감소세를 보이고 있습니다.

아미노산으로 구성된 단백질은 알파폴드의 인공지능으로 그 구조를 예측할 수 있으며, 알파폴드는 놀라운 예측력을 입증했습니다.

인공지능과 합성 생물학의 융합은 신약 개발 및 기타 분야에서 상당한 발전을 가져올 것으로 예상되며, 인류의 역사를 재편할 수 있는 잠재력을 가지고 있습니다.

영리를 목적으로 성경을 인쇄하려던 구텐베르크의 인쇄기가 의도치 않게 과학 혁명과 종교개혁을 촉발시켰습니다.

인공 지능과 합성 생물학 같은 현대 기술은 이제 드론 전쟁이나 바이러스 병원균의 인공적 생성 같은 메타 문제를 일으키고 있으며 재난을 일으킬 잠재력을 가지고 있습니다.

DNA 합성기와 개인 실험의 증가는 잠재적으로 새로운 바이러스와 강력한 무기를 만들 수 있는 중대한 위험을 초래할 수 있습니다.

유전자 편집과 AI 기술의 윤리적 딜레마

허젠쿠이는 비윤리적인 유전자 편집으로 전 세계적인 비난과 3년의 징역형을 선고받았으며, 그의 연구 결과는 쌍둥이의 후손에게까지 영향을 미쳤습니다.

인공지능의 뛰어난 작문 능력으로 그럴듯한 가짜 뉴스가 만들어지고, 딥러닝 기술로 동영상과 이미지를 그럴듯하게 조작할 수 있게 되었습니다.

인공지능과 합성생물학 같은 신흥 기술이 경제 성장과 주요 강대국 간의 경쟁에 큰 영향을 미치는 것으로 여겨지면서 인공지능을 규제하려는 유럽의 노력은 비판에 직면해 있습니다.

기술 규제에 대한 국제 협력의 필요성

기술 규제의 글로벌 과제를 해결하기 위해서는 실무에 초점을 맞춘 생산 부서의 비평가가 필수적입니다.

저자는 공격적인 성장이 균형 잡힌 발전 노력을 저해할 수 있으므로 국제 질서를 고려하는 것이 중요하다고 강조합니다.

첨단 소재와 기술 부품에 대한 수출 규제는 필요한 규제를 위한 시간을 벌기 위해 제안되는 반면, 합성 생물학 분야에서 강대국 간의 협력 가능성은 그 파괴적인 결과에 대한 우려를 불러일으킵니다.