Tencent Yuanbao has been connected to DeepSeek R1, and articles from WeChat public accounts can be found through online searches.

Today, Tencent has connected DeepSeek R1 to the grayscale search on WeChat . Let’s ask Yuanbao to start the online search:

This is a killer combination of national-level AI and national-level applications !

The 2025 AI competition just started, and WeChat is preparing to end the competition early .

This article talks about the impact it will have on us . As usual, there is an easter egg at the end of the article.

01 Top AI enters the homes of ordinary people

At the beginning of 2025, DeepSeek gave technology a shot in the arm, greatly shortening the gap with OpenAI and other manufacturers, which was originally more than a year .

After becoming a global hit, cloud vendors rushed to deploy DeepSeek on their own platforms in order to allow more people to use it and enjoy its benefits. Major product manufacturers and startups also rushed to introduce DeepSeek into their products.

Until today, Tencent, which has always been a latecomer, finally took action.

Tencent's direct introduction shows that it is extremely optimistic about the effectiveness of DeepSeek , because Mr. Ma once said that only mature AI models will be introduced into Tencent's first-level entrance .

The introduction of national-level AI into national-level applications means that top-level AI will become available to ordinary people .

The time when everyone can use top AI models has been brought forward by 1-2 years . On the surface, Tencent’s quick access is very unexpected.

Because Tencent itself has its own AI big model, Tencent Yuanbao is an AI application based on Tencent Hunyuan big model . In addition to generating text, Tencent has also open-sourced a series of models such as raw pictures, raw 3D and raw videos .

But if we think deeply, in today's AI era, whoever can provide a better user experience and truly solve user needs will be more likely to have the last laugh .

Win-win thinking and open source cooperation are the rules of this era.

02 Agent Era: Silent Moisture

“

Agent is an intelligent body. If you don’t know what it is yet, you can ask DeepSeek. I won’t explain the technology here, just look at the scenario interpretation below.

After the deployment and inference costs of the model are reduced, it becomes economically feasible to run multiple professional-level agents at the same time , which will accelerate the arrival of the Agent Era. The slogan that the industry has been shouting for a year last year, "2025 is the Agent Era", is becoming a reality.

Within the WeChat ecosystem, there is a large and unprecedented opportunity for smart WeChat mini-programs .

Looking further out, in the future, the first-level search entrances of Meituan Takeout, Didi Taxi, Ctrip Travel, etc. will all be connected to Agent .

Maybe one day this year, when you use them, you will find that all operations are idiotic with just one click .

And, based on its comprehensive understanding of you and data analysis, it may know you better than you know yourself .

03 AGI A Thousand Miles to Jiangling in One Day

“

AGI: Artificial general intelligence, which has extensive learning and adaptability capabilities like humans.

The AI ​​wave triggered by DeepSeek has completely overturned traditional views.

I saw someone compare the "DeepSeek January Revolution" to the "Watt Steam Engine January Storm", because in the history of science and technology, the definition of greatness is: being able to reduce the cost of a cutting-edge technology product and popularize it globally .

I totally agree.

The push from the AGI era is getting stronger and stronger . The originally generally estimated 10 years will most likely be shortened to 5 years, or even shorter.

For ordinary people, the requirements for mastering deep thinking ability are getting higher and higher . When AI can handle all generalized information, human thinking must be terrifyingly deep in order to better welcome the future AGI era.

end

“

I asked DeepSeek a question: "Why do ordinary people have to learn AI now?"

Here is his answer.

Because learning AI is like learning how to use a computer 20 years ago, or how to use a smartphone 10 years ago. If you don’t learn it now, you may be tricked even when ordering takeout.

Essential for pit prevention

When watching short videos, you can tell which ones are fake people and fake stories generated by AI .

When you receive a scam call, you can tell whether the voice is synthesized by AI.

When looking for a job, know which positions will soon be replaced by AI (such as basic customer service and simple clerical work).

Earning Opportunities

Work as an AI model for Taobao stores to save shooting fees (hourly wage 300+).

Helped the community supermarket design promotional posters using AI (monthly income from side job: $2,000).

Use AI to write short video scripts (efficiency increased by 5 times).

Life is a cheat

Use AI to plan routes and bargain with one click when traveling .

AI can be used to check errors in children’s homework and explain them .

Elderly people use AI to translate the professional terms used by doctors when seeing a doctor .

Don't be eliminated

• Workers who can train AI > Workers who don't use AI

• Bosses who use AI > Bosses who don’t use AI

• Elderly people who understand AI > Elderly people who don’t use AI

Don't learn now? That's OK, it's equivalent to:

• Others use calculators while you calculate by hand

• Others drive while you ride a horse

• When someone else is video calling you, you send a pigeon message

Anyway, I won't starve to death, but I will live a more tiring life: