In 2006, the Internet made everyone the protagonist of the time. Therefore, Time chose “You” as the annual figure.
In just 17 years, a major change in the times.
Today, the AI chat robot ChatGPT breaks out.
what does that mean?
AI has become its own protagonist, creation, writing code, reply email …….
picture
And this is just a beginning …
The AI military reserve competition is opening
Since 2014, with the proposal and iteration update of deep learning algorithms represented by the genera confrontation network (GAN), the generation AI has entered the new era.
Until 2022, AIGC entered the explosive development stage.
Google has released text image generation model Imgen, Parti, Muse; OpenAI has updated its own Dall-E 2; META’s text generation video model Make-A-Video, etc.
The birth of ChatGPT allowed people to return to the day of 2016 that the crazy artificial intelligence was against Li Shishi in 2016.
This fanaticism made those technology companies investing billions of dollars in the field of artificial intelligence caught off guard and caused a fierce arms competition in Silicon Valley.
In just a few weeks, Microsoft and Google have changed the entire corporate strategy to seize the opportunity brought by the latest artificial intelligence.
Microsoft invested billions of US dollars to OPENAI and announced that he would integrate the format AI into its OFFICE software and search engines.
In response, Google first announced the emergency state of entering the “red code”, and then released the AI chat robot Bard.
On February 7, Microsoft CEO Nadella challenged Google, “A competition starts today. We have to act and move fast.”
Not only that, even Wall Street made a positive response, and the analyst raised the stock of AI companies mentioned in the plan.
Investors have betted the format AI.
Because it may shake the market like Microsoft Windows 95 or the first generation of iPhone.
This crazy gold rush fever may also be catastrophic. The generation AI is like a bubble that is gradually expanded, maybe it breaks.
Because the enterprise is eager to improve the technical level and profit from it, the research on ensuring the safety of these tools is placed in a secondary position.
This is a story of the gold rush fever, and history will tell us what happened next.
Openai
In fact, the problems encountered by social media are also issues that need to be faced by generating artificial intelligence.
Over the years, the Artificial Intelligence Research Laboratory has been confidentiality of the copyright of these tools, while studying its potential danger, from error information, hatred speech to geopolitical crises.
And this conservative behavior is derived from “unpredictability of neural networks.”
The human brain inspired by neural networks is a computing paradigm based on modern artificial intelligence.
Traditional computer programming methods depend on accurate instruction sets to produce predictable results, and neural networks can teach themselves to identify models in data. The more data and computing these networks enter, the stronger their ability.
picture
In the early 2010s, Silicon Valley realized that neural networks are more promising to strong artificial intelligence compared with traditional programming.
But the problem is that early AI was very easy to imitate prejudice in its training data, input error information or hatred remarks.
When Microsoft launched in 2016, the chat robot TAY was launched for less than 24 hours, and it released the news that “Hitler is right and I hate Jews” on Twitter, and “feminists should be burned to death in hell.”
Even in 2020, the GPT-3 launched by OpenAI also has problems such as racial prejudice.
The artificial intelligence boom was truly driving around 2020. Several key breakthroughs in neural network design, increasing data, and science and technology companies are willing to pay for huge computing power to promote this prosperity.
It is embarrassing that most of the companies including Google, Meta and Openai are mostly unwilling to disclose their cutting -edge models.
In April 2022, OpenAI released the model Dall-E2 of the text to the image, which is limited to the list of credible users. Openai said the use of these users will help it understand and solve the prejudice retained from its training data.
Although as of July, 1 million users have used Dall-E, many researchers in more artificial intelligence communities are disappointed with OpenAI and other artificial intelligence companies.
In August 2022, a London startup named Stability AI released a tool Stable Disusion, a text to image, and opened to the public.
picture
For a while, Stable Diffuse quickly became a hot topic.
Millions of users are attracted by their ability to create art from scratch. According to the different prompts given by the user, this tool continues to output to obtain poisoning.
“2022 The Artificial Intelligence Report” investor Nathan Benaich said, “This tool opened the Pandora Box to generate AI. This shocked Openai and Google because they can use their closed tools around the world now.”
Openai followed up quickly and opened the door of Dall-E2. Immediately after November, ChatGPT was released to the public.
At the same time, Microsoft, Google and Meta are developing this fanaticism to the extreme.
In February, Google released Bard to compete with ChatGP.
At the same time, at the recent quarterly financial report conference call, CEO Xiaoza claimed that his goal was to make the company the leader of the field of artificial intelligence.
And this AI arms race is advancing globally. Domestic technology giants Baidu and Ali announced their launch of their own chat robots.
In this boom, the errors and hazards caused by technology have increased, and strong opposition has increased.
When Google Bard answered an error in the demonstration, the stock price fell. Not only that, Microsoft’s Bing is also easy to make mistakes.
DeepFakes is being used to harass people or spread error messages. Companies, including Stability AI, are facing lawsuits from artists and copyright owners. These people oppose that they use their works without permission to train AI models.
A survey of Time found that OpenAI’s outsourced Kenyan workers with less than $ 2 Kenya workers to review harmful content, including hatred speech and violence.
Although these problems are worrying, compared with the possible situation of the competition, they are insignificant.
At present, many of the choices made by large technology companies just reflect the choices they have made, which can be said to be a devastating chain reaction.
Social media is a Silicon Valley’s innovation that truly changes the world, bringing the first valuable lesson.
Its emergence is to make the connection between people, make the society healthier and the happiness of individuals.
But its failure is that technology companies have turned it into a “money -making machine” -the slowly distorting our news release to optimize participation; let us browse targeted advertisements.
At the same time, society has to deal with the influence of the next stage. The damaged news business, the increase of error information, and the rising youth mental health crisis …
It is not difficult to see that the integration of artificial intelligence and large technology products will also embark on the same path.
Google and Microsoft are most interested in how AI will make their search engines more valuable and perform Bard and British demonstration.
But Hugging Face’s chief ethicist Margaret Mitchell believes that search engines are “absolutely the worst ways” using generated artificial intelligence because it often errors. The actual advantage of AI like ChatGPT is marginalized, which is to stuff this technology into a money -making machine for technology giants.
If the search engine successfully integrates AI, this subtle change may destroy many companies that depend on search, whether it is advertising traffic or business recommendation.
Microsoft CEO Nadella said that the new version of Bing Search Engine will bring more traffic to publishers and advertisers, thereby bringing more income.
But just like artists’ resistance to artificial intelligence generation works, many media people are now worried that in the future, technology giants will eat the content of news websites without providing any returns.
The problem of how artificial intelligence companies make their project make money are also prominent.
At present, the AI -type generating tools can be used for free, in order to exclude competitors and get huge investment subsidies from venture capital companies.
ChatGPT climb on Time cover! In this competition, Openai gambling won
Although companies that have unsuccessful this strategy will spend funds, the companies that win the winning will eventually win the market advantage.
Openai is exactly the winner in this competition.
You know, the cost of talking to users with users is amazing. The cost of each chat costs is a little or more.
It is Microsoft’s “generous sacity” that allows Openai to have a huge competitive advantage.