On November 7, 2023, OpenAl's first developer conference was held as scheduled.
There was a lot of surprise content at the developer conference, and we found the live video of the whole process. Friends who have seen it can revisit it a few more times, and those who have not seen it can also watch it, and take this opportunity to understand what is happening in the world at the moment.
As one of the earthlings who started to contact ChatGPT at the end of last year, I have seen and heard a lot in the past few days. The most commonly heard and seen are "Everyone Can XXXX", "XXX Sentenced to "Death", "XXX Becomes Cannon Fodder", "XXX is Unemployed"...
BUT
Did the person who wrote this content really use his brain and have a little thought about OpenAI and ChatGPT?
Of course, we can also say that "existence is reasonable". The self-media KOLs who write these contents actually have a deeper purpose - "selling courses to monetize", "establishing and operating private domains", "such titles have traffic"...
Now, two days have passed since OpenAI's developer conference, and the heat has been replaced by "OpenAI server crash" and "OpenAI server hacker" (PS: I don't see the previous self-media reporting and commenting on this content). We also want to "pour cold water on this to calm down the earthlings" and talk about different observation angles and feelings.
First: GPTS only belongs to people who understand products and needs, not really "everyone"
What made many people "exclaim" the most at this developer conference was the launch of GPTS. But if you really know about the AI industry, you will know that the concept of GPTS is actually similar to an extension of the "knowledge base". And there was a similar concept at the iFLYTEK press conference a while ago, and they called it "Xinghuo Assistant". Underpinning the concept before these two models is DIFY, a platform that allows you to build Q&A conversations based on models and your own knowledge base.
If you look closely, you will know that in this operation, the role of Knowledge is an "accurate knowledge base", and the so-called "view all gpt4 background knowledge" plays the role of a pan-knowledge base.
Let me give you an example, for example, if you serve a client's daily press release writing, if you want to build a GPTS for copywriting, if you use GP4 for pan-search integration, the information may not be accurate. If you upload the content of the press release and the regular information of the past year or more in Knowledge, it can play a role in accurately producing the content of the manuscript.
Of course, ChatGPT has more powerful functions, it is not a simple "knowledge base", it can realize the link and call between apps, which is really felt in the on-site display. If you want to make an intuitive analogy, the closest thing may be the form of WeChat mini programs.
Therefore, this is not really something that "everyone" can build without thresholds.
First of all, not everyone can use magic to register an account, and secondly, not everyone has a "product concept". GPT is a "product", you need to have customer needs, scenario requirements, and you need to know how to implement requirements, rather than a simple Q&A conversation in general. Of course, if you just want to make a "leek-cutting" product, you can be when we don't say it. Because there are too many people on the earth, there are always people who want to make "leeks". In addition, the information changes too quickly, and we ourselves have become "leeks" in the next second.
Second: keep learning, love learning, and keep learning
To be honest, what amazed me the most at this conference was not the ability of the various technologies of the government. It's a clip of this old man appearing in front of the camera to express his feelings about OpenAI.
The appearance of this old man reminds me of the famous quote from "JoJo's Bizarre Adventure":
"The hymn of humanity is a hymn of courage, and the greatness of humanity is the greatness of courage."
We often lament that "technology is moving too fast and leaving older people behind", but I actually want to say that technology is leaving people who don't want to keep learning. Learning is difficult, but if there is a person who has the heart to learn, I think he will never be abandoned by the times.
So: please keep learning, love learning, continue to learn mentality, in the future, as human beings, we will have more challenges, and even really face the scene of human vs. AI in science fiction, and at that time, if we don't have the accumulation and mentality of learning, we may not even be able to pull out the network cable.
Third: AI at this stage does not have the ability to replace human work
The advent of AI will not replace human work. At least at this stage, we don't need to have such anxiety.
Why?
Because as far as OpenAI's DDOS attack is concerned, the experience of a large number of GPT4 users will be affected. It's like if you hire an employee with a salary of millions a year, but he gets sick and can't go to work these days, your business has a huge impact, and most importantly you don't know when he's going to get back healthy.
In fact, the so-called "cost reduction and efficiency increase" brought by AI is not simply the concept of reducing labor costs, but on the basis of eliminating a group of migrant workers who "fish in troubled waters" and "eat and wait for death", to further improve work efficiency and output efficiency. And all of this requires an "excellent and serious" worker to command and manage the AI, in order to truly achieve the goal and achieve the desired results.
It's like everyone can use AI tools such as gpt, mj, and sd to produce materials, but only excellent copywriters and designers can really produce valuable works that customers pay for. Don't be too superstitious about the things that are shown to you on the Internet to show you all kinds of great results, the really valuable things are still produced by human creativity.