Today's social media was flooded by an app called 'Lingguang'.


Understood, this is an AI application from Ant Group, positioned as a zero threshold full modal AI assistant for ordinary people, and can also generate interactive small applications in natural language within 30 seconds. This reminds me of the upcoming release of Gemini 3.0 this week, which is still in its early stages. With a one sentence generative operating system, it gives us an imaginative picture. What are the boundaries generated by AI in the future?


Since the beginning of this year, there has been a clear issue in the field of AI assistants: how to make AI's generative capabilities emerge.


From OpenAI and Google, to domestic companies like Doubao and Alibaba Qianwen, many new products have emerged. Nowadays, the ability of AI to help you create web pages and videos seems quite ordinary. Leading companies are not only trying to use concise and intelligent AI to assist? By replacing traditional applications with complex operations, we are exploring more possibilities.


I have decided to start from this standard and test the product emergence ability of the Lingguang APP.


The first step is to log in. The login page is very minimalist, with a catchy slogan: Make it complex, make it simple.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%

What is complex? How can we simplify it again?


01

Firstly, torture the 'Spiritual Light Dialogue'


Lingguang has three basic functions: Lingguang dialogue, Lingguang eye opening, and Lingguang flash application.


Dialogue as an AI assistant? A very basic yet crucial function.


I have been browsing dinosaur science popularization videos on Bilibili recently, and then I came up with a question that is very suitable for torturing the ability of spiritual light dialogue:


What is the difference between Tyrannosaurus Rex and Velociraptor?


The reaction speed of Lingguang is good. It starts outputting in two seconds and "hands over" in less than ten seconds.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


Comparing multiple products, Lingguang's answer has a characteristic: it is short in words but has strong summarizing ability.


In all kinds of Internet information, this kind of "? I really like the ability to refine degrees myself.


For example, in the previous question, comparing the Tyrannosaurus rex to a heavy tank and the velociraptor to a special forces unit is a very clever summary, which highly summarizes their body shape, biting force, social and solitary living habits.


In addition, the spiritual light will actively intersperse pictures and provide answers with both text and images.


I checked and found that this is one of the characteristics of spiritual dialogue, not only can it? Create illustrations to simplify abstract and complex knowledge, presenting it visually as images or animations.


So I continued to ask it to explain the Tyrannosaurus rex to me, and presented it in 3D. As a result, it not only provided the image, but also scientifically restored the details:


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


Unlike many other generation methods on the market, the dialogue between text and images in Lingguang is particularly suitable for rapid research and scientific enlightenment.


From a technical perspective, Lingguang is the first to achieve multimodal content output with full code generation, using code as the core generation method to build structured, personalized, and interactive answers in real time.


Only in this way can we achieve a combination of graphics and text, and even generate 3D models and maps.


Lingguang has opened up a new way of information exchange, no wonder netizens commented that 'Hangzhou has DeepSeek at the beginning of the year and Lingguang at the end of the year'.


I continued to ask Lingguang, 'I want to depart from Beijing this weekend and go hiking in Tiger Leaping Gorge. Can you help me plan a three-day and two night itinerary?' So I got this answer.


The route and itinerary it provides are basically on par with the itinerary of high-end tourist groups, accompanied by pictures and check-in points, and even taking into account various accommodation and dining details.


Even listed various fees and dress reminders.


Can it be said that it serves as an AI assistant? Its planning ability has reached a very detailed level. And all of these contents are based on a single command that can trigger a whole set of actions, which is very important.

02

Lingguang's' Kill move ': Generate application in 30 seconds


What other generation methods does AI have besides dialogue?


I think the application of Lingguang Flash has given the industry an answer.


According to Ant's official statement, it only takes 30 seconds for users to use the Lingguang Flash app? Becoming an interactive response?.


Based on my previous experience, building an app first requires holding several days and nights of meetings with product managers and design colleagues to obtain a prototype of the app. Then, modeling colleagues are asked to render the model, and front-end and back-end colleagues are hired to build the web framework. Finally, optimized code testing is added and launched, which may take dozens of people several months in the process.


Although AI coding is very powerful now, most AI coding models still require users to have coding abilities.


And now the spiritual light flashes? On the premise of zero code, it can be hand kneaded in 30 seconds. I tried to make it build a popular science application for the solar system.


After creating the application, I found that it could be changed and adjusted at any time, and it prompted me to add comet effects. I decisively agreed. It not only implements the code for you, but also provides you with free UI design.


It went very smoothly, and then I tried different small applications, such as designing a decompression game.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


After a short while, I successfully started playing by hand rubbing.


Not only that, but it can also generate an application for setting off various star shaped fireworks, an application to help workers relax their eyes, and even a low-fat diet for weight loss? Electronic wooden fish flashing response? , as well as translation tools and storybooks?.


When I started playing the Lingguang Flash app, I realized that my level of development for this app was less than one percent.


The reason why the playability is so high is because the Lingguang Flash application is a conversational, zero code AI coding application? ，? Do you need an IDE for coding? Do you need to leave the dialog box and enter? Programming tasks can complete an application.


Its speed, right? Are other low code products coming out soon? An order of magnitude. I tested it and found that the time required to generate an application can be controlled within approximately 30 seconds.


Unlike most small applications on the market, applications generated by Lingguang are not just static front-end pages, but lightweight applications that can directly call backend capabilities such as large models.


I tried it out and built a translation tool using Lingguang, which not only allows for real use but also enables real-time input and translation. If we give children a storybook flash app, it can write stories and read them out on its own, completely replacing the storytelling machine.


These applications have complete front-end and back-end logic, which actually broadens the achievable scene boundaries, allowing the Lingguang Flash application function to not only display results, but also interact with the outside world in real time.


In the later stage, it is difficult to imagine what kind of applications people will develop using the Lingguang Flash app, after all, the imagination of netizens is so big that the developers of Lingguang may not even think of it.


But is the spiritual light flashing? The emergence of AI is a very meaningful thing. Previously, there was a barrier between ordinary people and professional developers, and complex code and modeling abilities were difficult for ordinary people. However, in the era of AI, ordinary people can also use tools such as spiritual light? Rubbing application.


Not only full? Ordinary? Curious about development? The desire to create and share enables'?? 'to become a product manager. Is this kind of AI coding flattened? The attempt to promote AI equality is very meaningful.


03

The spiritual light opened its eyes and opened up a new world for explorers


In addition to the Light Dialogue and Flash app, there is another feature:


Lingguang's eyes open


This feature, located at the bottom of the main page and easily accessible, is a highly imaginative feature of the Lingguang APP.


I found during use that it is not just a simple scan or image recognition in a phone, but a built-in 'AGI camera'.


I opened my spiritual light and casually held it up towards the windowsill to take photos. Accidentally, it helped me solve the long-standing puzzle of what kind of hydroponic plant my colleague was raising on the windowsill.


Then, I held up my phone and experienced it in the office, making many unexpected discoveries, such as the brand and model of the heating device that my colleagues had been planting grass for a long time, the name of the figurine on the desk, and even the brand of the coat worn by the intern.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


In short, it gives me a feeling that when I open my spiritual eyes and look at any object, scene, or text in the real world, I will make some new discoveries.


Sometimes it can popularize the origin of a certain plant, a certain text, or a certain mechanical structure. When passing by a store and taking a picture of a sign, you can ask about its history and characteristics. It has a feeling that it will actively "read" everything you see in front of you.


The experience of "shooting and asking in real time" makes me feel that the feeling is very smooth.


The previous dialogue with the spiritual light requires you to actively communicate and ask questions, and when the spiritual light opens its eyes, it is like a tour guide, a teacher, leading you to explore everything you see.


This sense of exploration will bring users a completely new interactive experience. It can not only explain what this is, but also acquire extended knowledge such as "why" and "how.


In addition, when facing students or academic researchers, this feature can also achieve "fast research", even for lengthy documents, reports, or academic articles, it can quickly help you organize the core points and present complex concepts clearly to you.


In addition to "explaining everything in the world", there is another very good experience for Lingguang to open his eyes:? Created by the creative function.


When opening the spiritual light and looking at a certain scene, it not only recognizes it, but also has a "create" button. After clicking, it opens a new continent.


I turned the camera towards an alpaca doll and a pot of tiger tail orchids in the office.


I entered the command 'Create a video for the alpaca to start eating grass', and in less than five seconds, I received a video.


Can a single sentence be used for static photos? Note? Fresh and vibrant? Fate? For current large models, this kind of image generated text may not be a difficult task, but in the Lingguang APP, there is no need to import images or input spells. The threshold is very low, and coupled with a natural and smooth flow, users will love to use it.


In these imaginative scenarios, we realize that there are still too many possibilities for the emergence of AI products. (Leifeng Network)


Across the ocean, the upcoming Google next-generation AI model Gemini 3.0 is rumored to have imaginative gameplay features such as building a website in 5 seconds, constructing a Windows system with just one command, and cloning YouTube with just one click. Although Lingguang's performance is not as good as Google Gemini 3.0, they are all exploring in the same direction, exploring the boundaries of generation. The difference of Lingguang is that its threshold is almost zero, which makes it extremely easy for ordinary people to use the latest productivity.


Every new application that emerges from every product, including the Lingguang APP, every illustrated answer, and every Lingguang eye opening that explores the world, makes us realize that we are on the path to creating more possibilities for AI products.


Just as human development of the brain is less than 1%, our development of spiritual light apps is less than 1%, and we use spiritual light to achieve our spiritual light is also less than 1%.


At some point in the future, AI assistants like Lingguang will no longer exist as "tools" on the other end of the screen, but as "companions" exploring the world alongside you and me.

Today's social media was flooded by an app called 'Lingguang'.


Understood, this is an AI application from Ant Group, positioned as a zero threshold full modal AI assistant for ordinary people, and can also generate interactive small applications in natural language within 30 seconds. This reminds me of the upcoming release of Gemini 3.0 this week, which is still in its early stages. With a one sentence generative operating system, it gives us an imaginative picture. What are the boundaries generated by AI in the future?


Since the beginning of this year, there has been a clear issue in the field of AI assistants: how to make AI's generative capabilities emerge.


From OpenAI and Google, to domestic companies like Doubao and Alibaba Qianwen, many new products have emerged. Nowadays, the ability of AI to help you create web pages and videos seems quite ordinary. Leading companies are not only trying to use concise and intelligent AI to assist? By replacing traditional applications with complex operations, we are exploring more possibilities.


I have decided to start from this standard and test the product emergence ability of the Lingguang APP.


The first step is to log in. The login page is very minimalist, with a catchy slogan: Make it complex, make it simple.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%

What is complex? How can we simplify it again?


01

Firstly, torture the 'Spiritual Light Dialogue'


Lingguang has three basic functions: Lingguang dialogue, Lingguang eye opening, and Lingguang flash application.


Dialogue as an AI assistant? A very basic yet crucial function.


I have been browsing dinosaur science popularization videos on Bilibili recently, and then I came up with a question that is very suitable for torturing the ability of spiritual light dialogue:


What is the difference between Tyrannosaurus Rex and Velociraptor?


The reaction speed of Lingguang is good. It starts outputting in two seconds and "hands over" in less than ten seconds.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


Comparing multiple products, Lingguang's answer has a characteristic: it is short in words but has strong summarizing ability.


In all kinds of Internet information, this kind of "? I really like the ability to refine degrees myself.


For example, in the previous question, comparing the Tyrannosaurus rex to a heavy tank and the velociraptor to a special forces unit is a very clever summary, which highly summarizes their body shape, biting force, social and solitary living habits.


In addition, the spiritual light will actively intersperse pictures and provide answers with both text and images.


I checked and found that this is one of the characteristics of spiritual dialogue, not only can it? Create illustrations to simplify abstract and complex knowledge, presenting it visually as images or animations.


So I continued to ask it to explain the Tyrannosaurus rex to me, and presented it in 3D. As a result, it not only provided the image, but also scientifically restored the details:


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


Unlike many other generation methods on the market, the dialogue between text and images in Lingguang is particularly suitable for rapid research and scientific enlightenment.


From a technical perspective, Lingguang is the first to achieve multimodal content output with full code generation, using code as the core generation method to build structured, personalized, and interactive answers in real time.


Only in this way can we achieve a combination of graphics and text, and even generate 3D models and maps.


Lingguang has opened up a new way of information exchange, no wonder netizens commented that 'Hangzhou has DeepSeek at the beginning of the year and Lingguang at the end of the year'.


I continued to ask Lingguang, 'I want to depart from Beijing this weekend and go hiking in Tiger Leaping Gorge. Can you help me plan a three-day and two night itinerary?' So I got this answer.


The route and itinerary it provides are basically on par with the itinerary of high-end tourist groups, accompanied by pictures and check-in points, and even taking into account various accommodation and dining details.


Even listed various fees and dress reminders.


Can it be said that it serves as an AI assistant? Its planning ability has reached a very detailed level. And all of these contents are based on a single command that can trigger a whole set of actions, which is very important.

02

Lingguang's' Kill move ': Generate application in 30 seconds


What other generation methods does AI have besides dialogue?


I think the application of Lingguang Flash has given the industry an answer.


According to Ant's official statement, it only takes 30 seconds for users to use the Lingguang Flash app? Becoming an interactive response?.


Based on my previous experience, building an app first requires holding several days and nights of meetings with product managers and design colleagues to obtain a prototype of the app. Then, modeling colleagues are asked to render the model, and front-end and back-end colleagues are hired to build the web framework. Finally, optimized code testing is added and launched, which may take dozens of people several months in the process.


Although AI coding is very powerful now, most AI coding models still require users to have coding abilities.


And now the spiritual light flashes? On the premise of zero code, it can be hand kneaded in 30 seconds. I tried to make it build a popular science application for the solar system.


After creating the application, I found that it could be changed and adjusted at any time, and it prompted me to add comet effects. I decisively agreed. It not only implements the code for you, but also provides you with free UI design.


It went very smoothly, and then I tried different small applications, such as designing a decompression game.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


After a short while, I successfully started playing by hand rubbing.


Not only that, but it can also generate an application for setting off various star shaped fireworks, an application to help workers relax their eyes, and even a low-fat diet for weight loss? Electronic wooden fish flashing response? , as well as translation tools and storybooks?.


When I started playing the Lingguang Flash app, I realized that my level of development for this app was less than one percent.


The reason why the playability is so high is because the Lingguang Flash application is a conversational, zero code AI coding application? ，? Do you need an IDE for coding? Do you need to leave the dialog box and enter? Programming tasks can complete an application.


Its speed, right? Are other low code products coming out soon? An order of magnitude. I tested it and found that the time required to generate an application can be controlled within approximately 30 seconds.


Unlike most small applications on the market, applications generated by Lingguang are not just static front-end pages, but lightweight applications that can directly call backend capabilities such as large models.


I tried it out and built a translation tool using Lingguang, which not only allows for real use but also enables real-time input and translation. If we give children a storybook flash app, it can write stories and read them out on its own, completely replacing the storytelling machine.


These applications have complete front-end and back-end logic, which actually broadens the achievable scene boundaries, allowing the Lingguang Flash application function to not only display results, but also interact with the outside world in real time.


In the later stage, it is difficult to imagine what kind of applications people will develop using the Lingguang Flash app, after all, the imagination of netizens is so big that the developers of Lingguang may not even think of it.


But is the spiritual light flashing? The emergence of AI is a very meaningful thing. Previously, there was a barrier between ordinary people and professional developers, and complex code and modeling abilities were difficult for ordinary people. However, in the era of AI, ordinary people can also use tools such as spiritual light? Rubbing application.


Not only full? Ordinary? Curious about development? The desire to create and share enables'?? 'to become a product manager. Is this kind of AI coding flattened? The attempt to promote AI equality is very meaningful.


03

The spiritual light opened its eyes and opened up a new world for explorers


In addition to the Light Dialogue and Flash app, there is another feature:


Lingguang's eyes open


This feature, located at the bottom of the main page and easily accessible, is a highly imaginative feature of the Lingguang APP.


I found during use that it is not just a simple scan or image recognition in a phone, but a built-in 'AGI camera'.


I opened my spiritual light and casually held it up towards the windowsill to take photos. Accidentally, it helped me solve the long-standing puzzle of what kind of hydroponic plant my colleague was raising on the windowsill.


Then, I held up my phone and experienced it in the office, making many unexpected discoveries, such as the brand and model of the heating device that my colleagues had been planting grass for a long time, the name of the figurine on the desk, and even the brand of the coat worn by the intern.


After testing the 'spiritual light', I realized that human development of AI assistants is less than 1%


In short, it gives me a feeling that when I open my spiritual eyes and look at any object, scene, or text in the real world, I will make some new discoveries.


Sometimes it can popularize the origin of a certain plant, a certain text, or a certain mechanical structure. When passing by a store and taking a picture of a sign, you can ask about its history and characteristics. It has a feeling that it will actively "read" everything you see in front of you.


The experience of "shooting and asking in real time" makes me feel that the feeling is very smooth.


The previous dialogue with the spiritual light requires you to actively communicate and ask questions, and when the spiritual light opens its eyes, it is like a tour guide, a teacher, leading you to explore everything you see.


This sense of exploration will bring users a completely new interactive experience. It can not only explain what this is, but also acquire extended knowledge such as "why" and "how.


In addition, when facing students or academic researchers, this feature can also achieve "fast research", even for lengthy documents, reports, or academic articles, it can quickly help you organize the core points and present complex concepts clearly to you.


In addition to "explaining everything in the world", there is another very good experience for Lingguang to open his eyes:? Created by the creative function.


When opening the spiritual light and looking at a certain scene, it not only recognizes it, but also has a "create" button. After clicking, it opens a new continent.


I turned the camera towards an alpaca doll and a pot of tiger tail orchids in the office.


I entered the command 'Create a video for the alpaca to start eating grass', and in less than five seconds, I received a video.


Can a single sentence be used for static photos? Note? Fresh and vibrant? Fate? For current large models, this kind of image generated text may not be a difficult task, but in the Lingguang APP, there is no need to import images or input spells. The threshold is very low, and coupled with a natural and smooth flow, users will love to use it.


In these imaginative scenarios, we realize that there are still too many possibilities for the emergence of AI products. (Leifeng Network)


Across the ocean, the upcoming Google next-generation AI model Gemini 3.0 is rumored to have imaginative gameplay features such as building a website in 5 seconds, constructing a Windows system with just one command, and cloning YouTube with just one click. Although Lingguang's performance is not as good as Google Gemini 3.0, they are all exploring in the same direction, exploring the boundaries of generation. The difference of Lingguang is that its threshold is almost zero, which makes it extremely easy for ordinary people to use the latest productivity.


Every new application that emerges from every product, including the Lingguang APP, every illustrated answer, and every Lingguang eye opening that explores the world, makes us realize that we are on the path to creating more possibilities for AI products.


Just as human development of the brain is less than 1%, our development of spiritual light apps is less than 1%, and we use spiritual light to achieve our spiritual light is also less than 1%.


At some point in the future, AI assistants like Lingguang will no longer exist as "tools" on the other end of the screen, but as "companions" exploring the world alongside you and me.