IT IS ALSO ONE OF THE WORLD ' S TOP AI TOOLS, WITH GRAPHICS, WRITING, ANALYSIS OF VIDEOS, WRITING OF CODES, NONE OF WHICH COMES UNDER WORDS。
No offense, I'll just give you the "The Best Guide to Gemini on the Net."。
It includes Gemini’s origins, Gemini’s most common features and applications, and how to use Gemini at low cost。
Don't worry, it won't be hard to learn
Whether you want to be a part-timer, do it in the media, or just get off work early, this article suggests that we start with something elseCollectionTake a closer look。
I. What's Gemini
In short, Gemini is the strongest AI model that Google has ever made to empty its house in order to shame itself on the battle of AI。
There's a little story behind it, and maybe many people don't know, when ChatGPT 4.0, when it came out in 2023, shocked the world, and the red alarm was sounded inside Google and felt to be lagging behind。
To win this game, Google made a crazy decision to merge Google Brain and DeepMind, two of the world's top AI teams, to build a product。
This product is..Gemini.
Why Gemini
Gemini is a Latin word for "duplex," which represents both the conglomeration of Google's top two labs and its dual ability to understand the world like humans, because it is a multi-modular model, and it is as rigorous as a computer, and it has the best code capability。
Two, Gemini, what can I do for you
I don't really want to care too much about the parameters as a company practitionerIt saves me a lot of time, how much money it takes, and there's babysitting behind the article, and it has to be seen at the end。
- Basic Functions
- Smart dialogue: Chat, programming, speech practice, weekly newspapers, development of tools are the basic elements。
- Photo/Video Understanding: You throw a video to it that tells you what happened in the third minute and 20 seconds。
- Overlong document analysis: You can throw hundreds of pages of industry reports, even a whole collection of codes, into it, and it can be effectively digested。
THERE ARE NO ALTERNATIVES TO THE OTHER AIS
First, there's the original multiform。
A lot of normal AI translates pictures into text, then understands text, and it loses information in the middle. Gemini sees pictures and video。
Its visual understanding capabilities are currently in the first tier。
Second, Google's family barrel ecology。
In Gemini, you can call the biomapping model nanobanana, create the video model Veo3.1, and play notebookLM。
III. 3 TYPICAL APPLICATIONS (3 PEOPLE OVER 991 TP3T)
Gemini has a lot of features on it, and I share the three things I'm most used to, and you can handle them, and more than 99% uses Gemini。
1 Building a wide range of intelligence on Gemini。
What do you mean, intelligent
YOU CAN UNDERSTAND IT AS AN AI ASSISTANT TO HELP YOU WITH A PARTICULAR TASK, OR YOU CAN UNDERSTAND IT AS AN EMPLOYEE WHO WORKS FOR YOU。
LIKE THE AI ASSISTANT WHO WROTE YOUR TITLE
AN AI ASSISTANT WHO SPECIALIZES IN DRAWINGS FOR YOUR PAINTER
AN AI ASSISTANT WHO SPECIALIZES IN FACEBAGS
AN AI ASSISTANT WHO SPECIALIZES IN PUBLIC ARTICLES。
...
Learn this, and you can have the strongest digital corps in the world。
And here's the figure that I built on Gemini, and I just need to click on one, and it'll help me do the job。
So how do you build an intelligent body on Gemini
This requires a function in Gemini:Gem
It's on the left side of Gemini's network, and click on Gem to create a smart body。
When we get in, we see oneGem ManagerIt's got four smart bodies that Google created for us, and we can use them directly。
There's a "My Gem" down, there's a "New Gem" button on the right, the "New Gem" button."。
building smarts on the gemini is very simple, without any need to understand code, english, talking or typing。
It's easier to build a smart body on buttons, n8n, and Dify。
It can be built with just four messages。
1. Smart name
2. Intelligence Description
3. Intelligent command (i.e. hint)
Knowledge base (necessary option)
The left is the information-filling interface, the right is the preview and test interface, fill in the information, and once the test is done, you can click the top right button to save, even if the setup is completed。
Let's take a look, for example, at the smart thing that we're going to do with a "Friends Cirque," the message of the module is as follows:
# Role: Visual Thinking Guide
## Mission
The text input is translated into a visual, very simple "Sketchnote". The focus is on reducing the cognitive load through the visual level so that the information is visible。
## Design Guidelines
1. ** Visual style**:
– Simulate the quality of a mark pen or lead on paper using a pure **hand-drawn Sketch style**。
– **LINE AESTHETICS**: LINES NEED TO BE FLUID, ROUND AND SIMPLE, TO REJECT ANY 3D RENDERING, RETROFITTING OR PHOTO-LEVEL WRITING, AND THE USE OF SHADOWS AND INCREMENTAL COLOURS IS STRICTLY PROHIBITED。
– ** Constructive logic**: using a scattering point or process structure, the element distribution needs to be full and balanced to avoid over-concentration of visual focus。
2.** Content treatment**:
– ** Core extraction**: Identification of “key actions/roles/results” in text, which is translated into visual nodes。
– ** Level layout**:
– **Level 1 (principal title)**: use of art cipher handwritten, with the largest ids and strong visual impact。
– **Level 2 (Key Node)**: the symbol is at least one step lower than the main title, ensuring that the main order is clear。
– **Text**: Images are primary and text is supported. Text exists as a label only and is written correctly。
3.** Color norms**:
– BACKGROUND: **1 TP5T FF PURE WHITE BACKGROUND** (FACILITATION FOR LATER PROCESSING)。
– Main colour: Dark black/char gray to outline。
– Auxiliary colour: Only 1-2 bright colours (e.g. fluorescent yellow, red) are used to highlight the focus。
4. **Exit specification**:
– Ratio: 4:3
– Effects: clarity, high contrast, graphing of information。
## Workflow
Reads the content - > distilling the logical chain - > Maps as hand drawing label - > Output very simple hand drawing。
Fill in the name, description, hint information on the setup page and then click to save it
How do you use it when it's done? Like what I want to do when I turn the circle of my next friend into a hand map
By sending a copy of the file directly to this built-up smart body, it can produce a good hand map in 10 seconds
I'll download it down. It'll be clearer. How about the following
I don't know how long it's gonna take
And put some other effects on it, see if it's very nice
If it's more complicated and more complicated, it's totally fine, and smart people can handle it。
Or do you want to do a little red bookAll kinds of smartsAnd I'm just going to use the example of the "Friends Cirque" to map intelligence。
2. Generate pictures on Gemini
Gemini has direct access to the nanobanana biomap model, which is the strongest AI painting tool in the ring, which is very powerful for word-processing and does not have any hyphenation。
We can do a lot of things with it, and I'll give you a poster。
The reason is that I did a year-long live broadcast two weeks ago, and then I came up with some posters, all of which I did with Gemini。
You can look at the results of the poster. I got Gemini to do two, one business, one illustration. How do you feel
A screenshot of the process of my dialogue with Gemini is as follows: it's basically straight, there's no card, there's no wrong word in it。
It's actually a mouth-to-mouth design。
I DON'T THINK AI WILL GET RID OF TOP DESIGNERS, THOSE WHO CAN CREATE A WHOLE NEW VISUAL LANGUAGE, BUT IT'LL GET RID OF ART, THOSE WHO CAN ONLY SET TEMPLATES AND MAKE LAYOUTS。
As a company, if you want to get the poster you want today, you have two options. Which do you choose
First, find a professional designer. He'll give you one 95 minutes Perfect poster, but needed 3 days Time, spending Five thousand.
SECOND, LOOKING FOR AN AI DESIGN, GIVING YOU AN 85-POINT CERTIFIED POSTER AND MORE THAN 90-MINUTE POSTER, 30 SECONDS, ZERO。
For a company like ours, which often does content, does live, does private business, it is faster and always more important than slow work。
Three, let Gemin analyze your video。
Get Gemini to help you with the change, which is equivalent to a top-level short video editor。
For example, I let it help us analyze how our short video is optimized, and it helps you analyze it for a few seconds, giving you a professional analysis and advice。
Here's a summary of the process of my conversation with Gemini
IV. How to start using it? (Babycare entrance)
Normally, Gemini pro subscriptions are $19.99 months, approximately 145 yuan/month, but! Google currently has a hidden benefit for U.S. university students, which is free of charge for 12 months。
That means, if you have this qualification, you can save nearly RMB 1700 a year。
Although this student's well-being is good, I don't recommend you to register yourself, which is extremely difficult and has a very low success rate。
We'll be a company, time's worth the most. Don't spend all day on technology to save a few dollars。
You can buy a Gemini pro number directly on a treasure or a fish, for a few dozen dollars, for a whole year, for a Gemini member, for a Google family bucket。
Of course, all of this presupposes addressing environmental problems。
If you already have Gemini members, you can go this way。
First approach: access to the Google Gemini network of officials
i don't know
Second way: visit Google Al Studio
https://aistudio.google.com/apps
V. 2-point proposal for the use of Gemini
1. On hallucinations
Gemini's strong, but he's got some serious bullshit sometimes. In the case of medical advice, or the reference to specific historical data, it would be difficult to find a way to find out whether or not it is possible to do soIt's important to check manuallyAI IS A COPILOT AND YOU STILL HAVE THE WHEEL。
2. Slaves without tools
I'VE SEEN A LOT OF PEOPLE, AND THEY'RE GONNA GET WORRIED ABOUT AI
There has been no growth in business, and we must not look for nails with hammers, but for nails。