Today, we republish a popular article "2024, 7 Opportunities", hoping to inspire you.
Below is the text of this repost.
At the beginning of the new year, I always hear some friends ask questions like this:
Last year, already "just like that". How should I grasp the new year? What direction will the world change? Where will I still have opportunities?
This time last year, in order to answer these questions, I read with you a report called "Look at DAO2023", from unfinished research. There's a lot of talk about the outlook for technology trends in the new year.
For example, it talked about semiconductors at the end of 2022: "China's semiconductor industry needs a new round of industrial policies that keep pace with the times in 2023 and focus on manufacturing weaknesses." Talking about new energy: "China will account for half of the new renewable energy installed capacity in the next five years. ”
At the beginning of 2023, the big funds of the "national team" will be launched one after another, focusing on the layout and investment in the chip manufacturing industry. In mid-2023, the International Energy Agency (IEA) reported that China will account for 55% of the world's new renewable energy capacity this year and next.
So one after another, many readers left messages in the background saying that they were interesting and inspiring. So this year, I helped you find the new "Watch DAO2024".
In 2024, are there any changes that I don't necessarily know but are likely to land this year? Where will my life be affected? Are there any changes that may become opportunities for me?
With this question in mind, I finished watching this year's "Watching DAO2024". After reading it, I feel that the new changes, new trends and even new opportunities mentioned in it are particularly interesting and worth knowing as soon as possible.
For example, embodied intelligence. Spatial calculations. AI assistant. Gene editing. Autonomous driving. Small models. AI skits.
What are these things? How are there so many good-looking students?
It doesn't matter, all the new opportunities, at the beginning, only a few people can see, understand, and seize.
Today, I will share it with you before the Chinese New Year, so that the next time you see them, you will feel familiar and understand, and you will not miss them.
Let's start with a pair of glasses that excited a lot of people last weekend.
（
Spatial calculations
Last weekend, there was a pair of "glasses" that made many people so excited that they couldn't sleep.
On January 19, Apple's Vision Pro officially went on sale in the United States. The price of a pair is set at 25,000 yuan, which can be worth 3 iPhones, but the online server is still crowded, and it is sold out in an instant. The price of domestic scalpers has been speculated to 90,000 yuan.
Vision Pro? Why is it so expensive, and why is it so hot?
This pair of "glasses" looks like this:
What's more, after wearing it, your world may look like this:
What does this mean? What does it have to do with my 2024?
Probably, much more.
Back in 2023, when Vision Pro was just unveiled and released, Apple CEO Tim Cook couldn't hide his excitement and said this:
"Today marks the beginning of a new era in computing technology. Just as the Mac brought us into the era of personal computing, the iPhone brought us into the era of mobile computing, and the Apple Vision Pro brought us into the era of spatial computing. ”
In the future, whether you buy Vision Pro or not, play VR or AR games or not, the world you live in may begin to enter the era of spatial computing.
What is Spatial Computing?
It is one of the basic technologies of the metaverse and the next technological outlet in many people's hearts.
But most importantly, it's a technology that could revolutionize the way you interact with the internet:
It allows you to move from the era of "repeated horizontal jumps" in the real world and the digital world to the era of "deep integration".
Let's put it this way, in the past, whether it was the PC era or the mobile Internet era, whether it was a computer or a mobile phone, if you wanted to go online, did you have to face one thing: the screen?
In the digital world on the other side of the screen, you read novels, send pictures, watch live broadcasts... Mostly in 2D. In the real world on this side of the screen, you're living in a 3D world.
A digital world in 2D. The real world in 3D. These two worlds may occasionally mirror each other, but they are not in the same dimension in essence, how do you live?
Repeated horizontal jumps.
Walking, a few strokes of the circle of friends, you have to look up at the road. Eating, watching ten episodes of skits, turned around and found that it was dark. You have to choose between two worlds, going in and out, jumping sideways repeatedly.
Until, with the technology of spatial computing. In the era of repeated horizontal jumps, there is the possibility of being subverted.
To go online later, you may just need to put on your "glasses" and enter a space like this:
In this space, you can see two worlds at the same time.
On the one hand, you can see the furniture of the real world surrounding you, and on the other hand, you can also see the operation interface of the digital world running across the air.
Even, even the screen, and even the mouse and keyboard that were indispensable for human-computer interaction in the past... are no longer necessities.
In the future, with the blessing of spatial computing, you can control everything with a turn of your eyes, a gesture, and a command.
The real world and the digital world, at this moment, are blended and become one.
From repeated horizontal jumps to deep integration. From human-computer interaction that requires a mouse and touchscreen, to hands-free operation. From the two-dimensional world behind "one screen" to the three-dimensional world behind "a pair of glasses".
Behind the new era, there are new opportunities and new competition.
"Look at DAO2024" mentioned that the giants rarely mention the metaverse again. Instead, there is competition for spatial computing.
This competition is at a critical juncture. Next year is likely to be the "first year of spatial computing".
In 2023, Meta released the Quest3. In 2024, Apple released the Vision Pro.
The digital world is merging with the physical world to form a "new reality".
And when all this change happens, e-commerce, social, entertainment... A lot of things have new opportunities and deserve to be done all over again.
Spatial calculations of the track, the gun went off.
At the same time, another track is also on the way.
Embodied intelligence
Question: Do you remember last year's day, what was the hottest?
ChatGPT, definitely counts as one. I still remember that at the beginning of last year, ChatGPT was born, breaking the record for the fastest user growth and swiping screens all over the world.
The circle of friends is sighing, this AI chatbot is too powerful. It can actually chat with you like this, and can help you write so much, and even write code, exams...
However, a year has passed in the blink of an eye. The "hot words" of the market have quietly changed.
After watching a lot of AI chatbots like ChatGPT, many people have started talking about a new word this year: embodied intelligence.
At the ITF World 2023 semiconductor conference, NVIDIA CEO Jensen Huang said that the next wave of artificial intelligence will be embodied intelligence.
What is Embodied Intelligence?
The direct answer is: an intelligent system that can perceive, understand, reason, and interact with the physical world.
Translated, it is: like ChatGPT, it is also an AI robot, but in addition to having a "mind", it also has an additional "body".
What do you mean? I'll give you an example. It's all AI bots, and if you tell ChatGPT you're thirsty, it might say "drink some water" to you on the other end of the screen.
But if you tell the embodied intelligence that you are thirsty, not only will he be able to reply to you with a "drink some water", but he may also take a step to find the kettle himself, and then use his little hands to help you boil it, and finally pour the hot water into the cup and deliver it to you.
It can help you not only in the online digital world, but also in the offline physical world.
This also means that you must not only have the comprehension ability to accompany you to "chat" like ChatGPT, but also have a series of capabilities such as perception, understanding, decision-making, and control that can help you "take a hand".
It's amazing. What kind of robot can achieve so many capabilities at the same time?
"Seeing DAO2024" has an important point here: humanoid robots.
For example, like a human being, there are hands and feet.
The report mentions that over the past 20 years, the number of patent applications for humanoid robot technology has gradually increased, and the focus has gradually shifted from lower limb structure and gait control to arm structure and motion control, and there is a trend of further increase.
"Embodiment" can actually be very down-to-earth, for example, so that the robot has hands and feet.
For example, it is best to be close to people.
For example, the humanoid robots that will appear in 2023 are mostly between 1.5 and 1.8 meters tall. It looks friendly on the outside, and it can better integrate into the daily human environment.
Humanoid robots also pay attention to height and weight
And all this is to better serve humanity.
For example, they can understand "human words", transform "human words" into different subtasks, adapt to a variety of environments, and interact with the environment independently, so as to successfully perform a variety of tasks.
Such a powerful humanoid robot, will it be very expensive? "Look at DAO2024" gives a judgment: it is not expensive.
The report notes that generative AI is accelerating the growth of humanoid robots. Technology, market and policy work together to give birth to humanoid robots into the era of mass production, which are cheaper than cars.
What does this mean? New opportunities for a lot of people.
"Kan DAO2024" mentions two things in the report.
First, the market embraces humanoid robots.
Humanoid robots, tireless, without a demographic crisis, and in the long run, a lower cost per time. This year, Amazon trialed Digit, a humanoid robot, in its own logistics warehouse. Mass production of the Digit is scheduled for 2024.
Who is this opportunity for? Someone who can design a powerful agent system, someone who has the data needed for robot training...... But at the end of the day, it belongs to those who first saw it, and actively embraced humanoid robots.
Second, the government encourages humanoid robots.
China is the world's largest robotics market, and supply chain companies are increasing production capacity in China, according to the International Federation of Robotics (IFR).
Since 2023, governments at all levels in China have begun to guide entrepreneurs and investors to seize opportunities. From top-level design, to industrial funds, to policies and regulations.....
New tracks have opened.
When AI has a "body" and can help more and more, what opportunities will there be?
At the end of 2023, an elderly man answered this question with an annual open letter.
His name is Bill Gates.
AI Assistant
At the end of 2023, Bill Gates predicted such a scenario in a blog post:
"In the not-too-distant future, anyone who goes online will be able to have an AI-powered personal assistant that is far beyond today's level of technology. These agents will be able to help in almost any field of activity and life. ”
Having an AI assistant sounds good. However, it also sounds like a pie. Are there any AI assistants that have been put into shape and may really affect my life in 2024?
Bill Gates's blog post and the report of "Look at DAO2024" both put the spotlight in the same direction: healthcare.
Do you know how many health questions search engines get every day? There are 1 billion, Baidu, there are hundreds of millions.
Behind the hundreds of millions of questions asked every day is a health need that keeps hospitals overcrowded.
As a result, there are endless consultations and endless queues. The doctor's smile became more and more tired, and the words became shorter and shorter, and the patient even broke his leg, or even simply endured it, and there was no need to resolutely not go to the hospital.
Is there any way to reduce costs and increase efficiency in the healthcare industry, and become more scaled up to the benefit of all?
Of course.
Have you noticed that taking a film for testing, looking and smelling... Many healthcare operations are essentially doctors collecting and processing data.
For example, your ECG, which is image data, your blood test report, which is text data, and your family medical history, which is time series data.
Coincidentally. Collecting and processing data is precisely the strength of AI.
"Look at DAO2024" mentioned two points in the report.
First, the AI model has been applied to the medical industry. Look at the picture.
Since 2023, with the advent of powerful pre-trained large models, especially open-source models, medical large models have also accelerated their iteration. Among them, the mainland is also an important participant, and you can see large models such as Bian Que and Materia Medica in the picture.
These large models, like the doctor's version of ChatGPT, can not only talk to you about your condition, but also say professional or popular words depending on the person you chat with.
So in 2024, will doctors be replaced by AI?
It's not that easy. Usually AI bots chat, occasionally hallucinate, give an inappropriate answer, most of them are acceptable, but medical treatment is not.
Medical care is a matter of human life, and you can't make a mistake.
Therefore, although the report mentions that some medical AI assistants, such as Google's Med-PaLM 2, have basically approached the level of "expert" in medical exams, with an accuracy rate of 85%, they can still only be used as "assistants".
"Kan DAO2024" judged that in the short term, the application scenario of generative AI used in medical services, with the least regulatory resistance and the highest certainty, is to play the role of a doctor's "clerical assistant".
What can this "clerical assistant" do? You can still look at the picture.
A sentence of "why is it bad?", you have to ask it 100 times a day? Give it to AI, which will automatically record it for doctors and then transcribe it into electronic medical records.
After asking, you have to ask "Have you ever had this situation before?", and ask it again 100 times, and hand it over to the AI to automatically help the doctor search for the patient's past medical history and examination results.
Finally, after reading it, as soon as the nurse who took over came, she had to say it again? Or can you hand it over to the AI, let the AI and the nurse align the precautions, and at the same time tell the patient to take the medicine on time.
Perhaps, this is what AI assistants are really about:
Don't replace a job, take away only the most repetitive, boring, and boring parts of that job.
To this end, the tech giants have launched a full-scale competition.
Amazon has launched HealthScribe, which automatically generates medical records, and Google is testing similar generative AI tools, according to DAO2024. In addition, in the mainland, Baidu's spiritual medical model and Tencent's medical model are also revolving around scenarios such as medical records.
The doctor's consultation may change in the future.
Of course, the medical scenarios that can be subverted by AI are not limited to consultations.
Gene editing
"Kan DAO2024" proposed that by the end of 2023, the world's first gene editing drug will be approved abroad, and the era of gene editing drugs has officially arrived.
The matter of "taking medicine to cure the disease" may also become different from what many people think.
What do you mean? I'll give you an example.
Before, what happened when you took the medicine? Stimulate your metabolites, neurotransmitters, hormones... Let them work well and get you back to health.
In the future, when the medicine is taken in, it may skip these "screws" of the executive layer in the body and go directly to find and deal with the highest layer: genes.
For example, if a person is anemic, then as soon as the drug enters the body, instead of entering the hematopoietic department to say hello and engage in relationships, let them cheer up and make more blood, but directly rush into the chairman's office to rummage through his computer, find the gene responsible for hematopoiesis, and see if there is a mistake in the program code in the gene, resulting in genetic defects and ischemia.
If so, knock out the wrong code, modify the DNA sequence, and set things right from the source, so that the hematopoietic team below can restore morale.
This is gene editing.
It sounds like a change of life. Isn't that what God can do? Can humans do it?
Previously it couldn't. But in 2012, two scientists discovered that a technology called CRISPR could help humans equip themselves with "genetic scissor hands" and "God's scalpel" to precisely find and cut any gene of any species.
This discovery won them the Nobel Prize in 2020, and several scientists who smelled the opportunity quickly set up a company in the next step, attracting investment, finding scenarios, and making applications to meet needs.
For example, it can be applied to the treatment of diseases and even the improvement of crops.
It's amazing. However, it is also very difficult to do this kind of thing that changes one's life against the sky.
Because, the human genome contains more than 3 billion base pairs. Sequencing and analyzing these DNA, and even precise cutting, is a very complex and huge computational problem.
It just so happens that this kind of question, AI will. Not only will he be good at it, but he will be good at it.
DNA sequencing, which used to take 10 years to complete, can be done in a day by letting AI do it.
Kan DAO2024 mentioned that artificial intelligence has accelerated gene editing and related clinical research.
More gene-editing drugs have entered clinical trials.
Many diseases that were previously considered incurable, such as heart disease and Alzheimer's disease, have new treatment possibilities.
See a doctor, take medicine. It's all about human life.
Equally life-threatening is autonomous driving.
Autonomous driving
Autonomous driving may also be changed by AI.
In the first week of 2024, there is a hot search, you may remember:
BYD's deliveries in the fourth quarter surpassed Tesla in one fell swoop and became the world's top electric vehicle sales.
It's amazing. What does Tesla think about this? Musk replied with a sentence:
"To a lot of people, Tesla looks like a car company, but it's an AI/robotics company. ”
What do you mean? Isn't it competing for car sales, and why do you emphasize that you are an AI company?
The answer comes back to autonomous driving.
What is Autonomous Driving?
Although there are many classifications and definitions of how automatic driving is, the first reaction of most people when they hear this concept is that it is 100% automatic:
Let people liberate and let the car drive itself.
But do you really dare to let go? Why do you dare to let go? Human lives are at stake, so just hand them all over to a car?
There are two ways to solve this problem.
One is to use rules and manage the car.
Cameras, radars, and sensors are installed. The main one is deaf and clear-sighted.
Then, let the car be recognized by the signals transmitted by the sensors, and then make decisions according to the rules set in the code, and drive like the most accurate machine.
The other is to use AI to teach the car well.
It should also be loaded, and it should also be deaf and clear-sighted, but at the same time, it must also have its own "experience" to ensure that in bad weather, poor lighting, etc., the car can also drive well like an "old driver".
How to turn a machine into a human "old driver"? Use the power of AI to make it learn well.
For example, give it a mind with a "neural network" like a human, equip it with enough computing power, and give it enough lessons.
"Kan DAO2024" mentioned that the AI model has changed the competitive landscape of the autonomous driving technology route. It's teaching electric cars to drive like a five-star driver.
For example, Musk's new version of fully self-driving technology, FSD V12, has begun to use billions of frames of human driving videos to teach himself how to drive.
Instead of writing hundreds of thousands of lines of code to help machines make decisions, hoard hundreds of thousands of chips to help machine learning make decisions.
Instead of piling up enough code and writing enough rules to ensure security, we use V to make the machine "learn" enough data and emerge enough "experience" to ensure security.
This is Tesla's sentence: We are AI. We're different.
Also saying something similar this year, there is another person: Fu Sheng.
Small model
Last weekend, a lot of my friends sent me a link to the live stream, saying:
Take a look. Fu Sheng and Zhou Hongyi are actually on the same stage.
What's going on? What could make them both on the same stage?
Click on it and take a look, the name of the live broadcast is: "How to innovate enterprises under the tide of AI? From technology carnival to enterprise landing - privatization model is king"
During the event, Fu Sheng talked about two words: "anxiety" and "opportunity". Both words, both about the big model.
First of all, what is a large model?
From the rise of the large model of ChatGPT in 2023 to today, the "bigness" of the large model has at least 4 layers.
The first layer of "big" is in the number of parameters: hundreds of billions of parameters.
And the "bigness" of the parameters brings the "greatness" of the ability. Learn, generate... Everything is strong, astronomy, geography... Understands everything.
With it, it's like having a super employee with 100 PhDs in hand.
How much benefit can such a strong employee bring?
Microsoft, which has invested in openAI, will turn over in market value in 2023, surpassing Apple to return to the world's first. Midjourney, which is engaged in the large model of pictures, has more than 40 employees and a valuation of $10 billion.
A big model is indeed a big opportunity.
However, in the face of this big opportunity that can subvert the times, many people may be anxious, not knowing how to keep up, and not knowing whether they will be left behind.
Because, this big opportunity, not everyone can benefit from it at the beginning. Behind the big opportunity, there is also a big cost.
To "train" a super employee with 100 PhDs, it is necessary to feed a large amount of data, invest a lot of computing power, and even pay hundreds of millions of dollars in electricity bills.
If you want to do a business with large models, it's not that the giants can't even get tickets. If you want to do business with a large model, you have to hire this super employee and use it to help you improve your efficiency, and the cost will not be low.
So what should we do? Is the opportunity of a large model destined to be only an opportunity for a few? Not necessarily.
Fu Sheng gave an example, saying that in 2023, the annual privatization authorization fee of a 100 billion parameter large model will be tens of millions. A customer found him and said that we wanted to be a customer service, but this AI model is so good, I can't invest tens of millions of dollars a year.
Is there a cheaper option without sacrificing performance?
Yes. Perhaps, you can change the "big" model into "small".
What do you mean?
Behind the big model, there are big parameters, big opportunities, and big costs.
However, today, your large model has hundreds of billions of parameters, and you are a master of general practice with 100 doctorates, but I am in real estate, so I want to hire a doctor who understands real estate, okay?
A single doctor with a professional counterpart, and a general doctor are more than an order of magnitude cheaper than the annual salary, but the KPI is completed, and the performance is not necessarily bad.
Similarly, a model with 10 billion parameters that has been specially trained and debugged is not necessarily worse than a model with 100 billion parameters.
Even, it will be cheaper, safer and more flexible.
In the wave of AI large models, there is not only an opportunity to "become bigger".
In 2024, making the model "smaller" and using fewer parameters and resources to meet a more precise demand can also create value that benefits everyone.
This is what Fu Sheng said, and he saw the opportunity: the "privatization" model.
This is also an opportunity for the "small" model mentioned in "Look at the DAO2024".
"Look at DAO2024" wrote: 2023 is the year of large models, and 2024 will be the year of "small" models.
More small models with billions to tens of billions of parameters, through innovations in model architecture, algorithms, training, and fine-tuning, as well as combined with external retrieval, the performance can be called tens of billions of parameters, and even equal to GPT3.5 (175 billion parameters).
However, if you're not in the industry and don't care about reducing costs and increasing efficiency, that's okay.
Do you usually watch dramas?
AI skits
In 2023, the skit will explode.
Crossing, rebirth, counterattack, revenge... Every minute can pinch people's hearts and step on the "cool point".
There are more and more people watching skits, and the industry that makes money from skits is becoming more and more popular.
Volume content, volume production, also volume cost, volume efficiency...
Someone thought of it, with AI volume.
That's right, in 2024, AI has already gone from text to video.
At the beginning of 2023, when many people mention AI, they still think of chatting with text, searching for things, and writing articles like ChatGPT.
Later, the explosion of Midjourney and Wenxin Yiyan made people discover that AI can not only process text, but also create pictures.
At the end of the year, along with short dramas, there is also the application of AI in video generation.
For example, using AI to swap faces. And it is not only used to replace the star face in the TV series, but also to replace the Asian face in the short drama. After changing, it went directly to the sea, and directly became a "new drama" aimed at the overseas market with all the protagonists "European face".
For example, using AI to write scripts. Isn't it just a counterattack? Isn't it just rebirth? Set a few keywords, and AI screenwriters can be generated in batches, all year round.
Another example is using AI to help with editing. Text quick cut, intelligent frame filling, intelligent rough editing...... You can even turn it into a dynamic brush, and you can swipe it on any image to make it move and turn it into a video.
It's amazing. And, it can be even better.
In December 2023, my circle of friends was swiped by a girl.
Click on it to see that a young girl in her 20s dropped out of Stanford to start a business and opened a company called Pika, which attracted most of Silicon Valley within half a year, raised $55 million, and reached a valuation of 200 million.
Well, it's amazing to be young and promising. But after swiping the screen again, I realized that the focus is still behind.
This company can actually use AI to directly generate videos.
For example, enter a piece of text, a picture... It can automatically generate or adjust a video, and the texture can be comparable to that of a movie.
Like this.
Pika Official Showcase
Such.
Pika Official Showcase
Or so.
Pika Official Showcase
Hey, no wonder, some people say that the evolution of AI is calculated in days. Iterative by day, thousands of miles a day.
"Watch DAO2024" proposes that in 2024, a large number of film and television dramas generated by generative AI will appear, impacting the film and television industry.
In 2023, many technologies will achieve the effect of generating images and videos in real time, and the impact on the content entertainment industry such as film and television, music, and games will be disruptive.
Moreover, the application of this area is not much in the giants at present.
There is still a chance.
Embodied intelligence. Spatial calculations. AI assistant. Gene editing. Autonomous driving. Small models. AI skits.
It sounds like there are a lot of opportunities hidden in every direction.
Who can seize such an opportunity? Let's take a look at the picture again.
If the competition in the AI era relies on the use of "computing power" to wrestle the wrist, then the GPU chip that can be used for AI acceleration such as the "Nvidia H100" is the necessary "muscle" for the wrist.
And the "muscle" reserves can be squeezed into every company in this chart, such as Meta, Microsoft, Google, Amazon, Tencent, Alibaba... All have a common alias: tech giants.
What is a tech giant? Is it just a market capitalization of more than a trillion dollars? After reading "Watching DAO2024", you may have some more specific feelings:
You think it's just money, and it's stockpiling more chips than the national stockpiles of many medium-sized countries, but it's actually in control of the data.
Not only is a huge amount of business data generated from its own business every day, but it has also invested heavily in data centers around the world.
For example, Microsoft's annual investment in data centers can easily exceed $50 billion, which is equivalent to the investment of a technology power in AI infrastructure.
You think that capital, chips, and data are all its advantages, but in fact, these advantages will also give rise to a new advantage: talent. The world's best talent.
So, the source of the world's most high-quality AI papers is not the world's highest-ranked universities, but Google.
Capital, chips, data, talent. Layer upon layer, complement each other, and jointly achieve the myth of computing power of technology giants.
Is the opportunity in 2024, and even in the entire AI era, destined to belong only to tech giants? Do start-ups still have a chance? Do ordinary people still have a chance?
I'm also curious about this question. I asked Kevin Kelly in 2023 for advice. His answer was: Yes.
"I don't think the giants of the AI era will emerge from the tech giants that already exist," he said. On the contrary, it may be in its infancy and little known now. ”
Christensen, a professor at Harvard Business School, also pointed out in "The Innovator's Dilemma", which made him the "father of innovation", that many small companies that lack resources and have no scale advantage still have a chance to win in the face of large enterprises that seem to be leading in all aspects.
And these coincide with some of the views in "Look at DAO2024", the report reminds:
Disruptive innovation, which often happens in startups.
Large companies have no shortage of good ideas and good resources, but in terms of vitality and speed, they are not as good as many small startups.
For example, the place that produces the most high-quality AI papers in the world is Google. But it was OpenAI that first used these theories to make a good model.
In the era of AI, start-ups still have opportunities, ordinary people still have opportunities, and we still have opportunities.
The year 2023 that has just passed is called the "first year of AI" by many people.
Bill Gates wrote this sentence in his blog at the end of 2023:
"This year is the first time I've used AI for work, not just as a novelty. I suspect a lot of people do. We are now at the beginning of a huge technological transformation. ”
That's true for Bill Gates, and so are many of us ordinary people. Many people have also used AI for the first time, learned about AI for the first time, and even heard about AI for the first time in the past year.
This may be why "Kan DAO2024" actually changed the style of previous years when looking forward to 2024 and chose to focus all the pages on one theme: AI.