The worst AI ad of the year has appeared. What is more terrible than aesthetic degradation is "not being human"
2024-08-14
한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina
The "most relaxed" Olympic Games has come to an end, with laughter and tears, victories and losses, but the biggest loser must be Google.
During the Olympics, Google made a TV commercial for Gemini, showcasing Gemini's writing abilities -- unfortunately, it wasn't used in the right way.
In the ad, a father uses Gemini to write a fan letter for his daughter to Sydney McLaughlin-Levrone, a famous American track and field athlete, expressing his daughter's love for her and her dream of surpassing her.
As soon as the ad came out, it was criticized so much that the comment section was closed. Alexandra Petry, a columnist for the Washington Post, was the most harsh in her criticism, shouting between the lines: I beg for a pair of eyes that have not seen this ad.
Although everyone may have different understandings of the same content, Google does actively suggest in this ad that Gemini can be as good as humans.
This is calledThe biggest minefield in AI marketing at the moment.
Microsoft also advertised its AI products during the Olympics, and highlighted how Copilot can assist users in their daily lives.
From streamlining itineraries to making plans, Coplit ultimately helps users achieve the "You empowered" effect, which is a completely different tone from Google.
Even Apple, which was compared with Google for its failure, did not repeat the same mistake this time. It was still an ad made for the iPad, but it used street billboards that were closer to life, and collaborated with a French illustrator to take a lively and cute route.
Apple has a history of this. In May this year, in the advertisement "Crush" for the new iPad, a huge hydraulic press crushed a piano, paint, and a game console, and also crushed the audience's glasses, which triggered a huge wave of criticism.
Despite Apple's experience, Google didn't realize that there was something wrong with its own ads. Perhaps it was because it was facing a lot of problems at the same time, or it may be because it didn't think it had done anything wrong.
Light of Hope
Things haven't been going well for Google lately.
On August 5, the U.S. District Court for the District of Columbia in Washington ruled that Google had illegally monopolized the online search market and violated U.S. antitrust laws. On the same day, the stock price of Google's parent company Alphabet fell by more than 5%.
This is the largest technology antitrust case since the antitrust showdown between the US Department of Justice and Microsoft at the beginning of this century. It is also another antitrust fine for Google this year. In March, French regulators issued a fine of 220 million euros.
The U.S. Department of Justice’s lawsuit reads: “Twenty years ago, when the Internet was just emerging, Google went from being a hastily established startup to becoming the darling of Silicon Valley with its innovative search methods.The Google of the past is no longer the company it once was. Today, Google is just the gatekeeper to an internet monopoly.」
In just a few lines, it tells the story of the ups and downs of a technology giant over the past 26 years, as well as the changing mental journey of the world in dealing with technological change.
Back in the first decade of this century, Google was the representative of openness and innovation in the last wave of the Internet. Those who have been online for a while may still remember the nickname "Gu Renxi", which was given by Google fans, meaning that Google is the light of hope for mankind.
The world's love for Google reached its peak in 2012. That year, Google Glass was released, and Sergey Brin took the audience on stage and took them through Google Glass to the perspective of a skydiver in the sky.
It flew all the way over San Francisco and landed on the roof of the exhibition center where the press conference was held; it then climbed up the wall and dropped straight down from the facade of the building, and finally a mountain biker rode his bike into the venue and onto the podium, completing the launch of this product.
That was a significant milestone in the history of science and technology. Today, when we look back at those video records, the images are shaky and the pixels are poor, but we can still glimpse them.In the last decade, when new technology "landed", everyone was excited and cheered.
That year, it had been 13 years since Google insisted on the slogan "Don't be evil". However, it was only three years before it reorganized Alphabet. After the company's reorganization in 2015, the slogan was rewritten; after 2018, it gradually faded out of the arena.
Alphabet's reorganization at that time certainly had many commercial considerations. However, at that time, American companies generally recruited and strengthened themselves, with such a vision behind them:Maximizing profits and shareholder interests and allowing society to distribute welfare on its own is precisely a manifestation of the pursuit of public interests.。
However, this vision was not realized. The reorganization and expansion ultimately led to monopoly, a completely different direction from the early Silicon Valley hacker culture. Google went from "don't be evil" to "don't be human".
What’s more, the iteration of technology may be faster than the technology people themselves estimate. Whether it is Google Glass or the search that Google has always relied on, in the eyes of users, it was once a more natural and harmless "progress" paradigm.
Search engines serve as the key to the world of information and provide retrieval services; mobile electronic devices such as mobile phones and glasses are the way for the digital world and offline life to interact with each other - they are just auxiliary to make life better, and never aim to "replace" people.
However, AI is not all the same.
What artificial intelligence is pursuing is automation. Whether it is through algorithms or machines, it is aimed at those jobs that are carried out in a predictable environment and in a prescribed manner. AI products based on large language models further broaden this definition, lower the threshold of the word "predictable", and once again improve the level of automation.
Technological elites are generally enthusiastic about automation, but working people view automation in a completely different light.
Automation also has basic laws
When talking about automation, the Luddite movement in the early 19th century is often paraded through the streets.
This movement against automatic looms was eventually suppressed by the British government, and in the end, the opponents who smashed the machines failed to stop the process of industrial automation.
This example is often used as an argument to prove that the wheel of technological progress is rolling and unstoppable. But to be honest, the economic form of the 19th century is really very different from today.
In the modern economy, the proportion of service and knowledge-based jobs is much greater than in the industrial eraThe impact of AI on these fields is completely different from the impact on manufacturing in the Luddite era. It will not only affect blue-collar workers, but also white-collar jobs, including professional and managerial positions. It may also intervene in people's consciousness through natural language - these did not exist during the Luddite movement.
Has automation replaced humans in modern times? Yes, it has.
In the 1920s, when telephone calls needed to be switched, this was done manually, mostly by young women. At the time, AT&T was the largest employer of telephone operators in the United States.
Over the next three decades, the automated telephone switching system gradually replaced human operators. However, it did not cause a major blow to the employment rate. This was because different trade unions took quick action to fight for the rights of workers, especially to ensure that new jobs brought about by new technologies could take over the previous labor force.
This is the biggest difference from the industrial age of the 19th century:Modern society promises people that differences can be coordinated through systems and rules so that they can adapt to each other.。
No matter how harshly Google is criticized, no one will sneak into the computer room and unplug the power cord.
As long as both sides face their respective demands, there is no need to go to the point of smashing machines. In 1962, at the most important stage of the automation transformation in the United States, a report issued by the Labor Management Policy Advisory Committee was delivered to the desk of then-US President Kennedy.
The cover letter lists three main points:
The committee agreed on the following key points:
1. Automation and technological advancement are vital to the country's overall welfare, economic strength, and national defense.
2. Such progress can and must be achieved without sacrificing human values.
3. Achieving technological progress without sacrificing human values requires cooperation between the private sector and the government based on the principles of a free society.
Kennedy wrote back, “I am very grateful for the contribution this report represents. The problem of automation is as important as any we face. We must take advantage of every opportunity of technological development, but we cannot lose sight of the human values involved.”
Imperfection breeds perfection
What exactly is "humanistic value"? Everyone has their own definition, but what cannot be escaped is the existence of "human beings" themselves.
The biggest mistake the Gemini ad made was to send out the subtext that AI could replace a person — and a child at that.
Children are the hope of society, but at the same time, they are definitely imperfect. Gemini can write a letter with smooth and flawless writing, but a child may even make mistakes in spelling words.
But is perfection the only thing worth existing?
Weibo celebrity Lan Xi commented, "The progressive nature of technology itself easily indulges in the growth of self-obsession, and the end point of self-obsession is arrogance." This is a very polite statement, as he did not clearly point out what the obsession and arrogance are.
It is precisely a perfect appearance.
When the automatic wiring system appeared, it was not all done in one go, and those thirty years were not entirely spent on coordinating labor. In the early days, the wiring system relied on mechanical components such as motors and relays, and required precise matching in a large number of lines, as well as challenges such as signal clarity and noise reduction, which made the reliability and stability of the system a big problem.
▲Around the 1940s, a telephone engineer was checking the connection of the telephone line. Image from: Georgia State University Library Archives
The whole society accompanies manufacturers and developers, adapting little by littleanswer.There were also criticisms during this process, accusing the machine of having rigid programs and being inhumane, but it was through such adjustments that the iteration process was finally completed.
Today’s generative artificial intelligence also has many problems. Card drawing and illusion, limited specific application scenarios, and the invasion of data and content are always challenging people’s nerves. However, if we really want to follow the perfectionist approach and have no tolerance for early and immature technologies, GPT would have died in the cradle when it was still 2.0.
Just like facing children, even though they can't write perfect words, you still have to let them write.
Google used the user data collected over decades to train its own "perfect" tools, and then turned around and said that human-produced products were not good enough and should be eliminated. This is simply the opposite of what is going on.
There are many invisible costs behind perfection. If you only fall in love with its smooth and white appearance, you will be easily deceived, indulge in arrogance, and be complacent about your achievements, but forget:It is the public's tolerance of all the early imperfections that leads to later perfection.。
Artificial intelligence as a whole has to face many more problems than a Google ad that failed. This episode reminds everyone that a coordination process based on humanistic values will be a long journey.
