It starts with innovation and rises with precipitation. The hard-tech industries waiting for the future have ushered in many “inflection moments” in 2019. Some have ushered in the first year of development. Some are making a lot of money while others are rushing rationally in a blend of ice and fire. For example, 5G and chips have become the blue ocean market that the industry must meet. At the same time, many companies in the field of autonomous driving and artificial intelligence are still staggering in the waves.
In the hot and dry summer of San Francisco, the local people gathered in groups in the streets, the sign with a thick red × was held high above their heads, and their mouths shouted “No Face Recognition!”
Google, Facebook, Amazon … As “Bay Area Light” technology companies were surrounded by the next floor, they have never been so turbulent as this year because of the negative impact of leaking user data privacy Dense public opinion crisis.
China on the other side of the Pacific is also trying to escape from technology companies.
The challenge of using AI video surveillance to enter the classroom has not completely subsided, and the face-changing app “ZAO” was interviewed because of illegal collection of face information, and then a consumer sued “Human in Hangzhou Wildlife World”. The first case of “Face Recognition”, one wave after another, one after another.
The endless stream of user data privacy leaks has made the contradiction between the development of cutting-edge technology and the protection of user privacy never as dazzling, sharp, and difficult to reconcile as this year. Technology has never been so strongly opposed by users.
Apologize, apologize, thank you behind closed doors. But apologies have not solved the problem, and the former “do no evil” creed has also proven to be invalid. After four or five years of high development, the sudden and concentrated negative events erupted the technology companies that have full control over user data and experienced an unforgettable crisis of trust.
The emergence of the crisis has also pressed the pause button for the development of the entire industry. While busy occupying the technical high ground, promoting the landing of commercialization, seizing the market, and capturing profits, technology companies and business elites have to start thinking about users as Human privacy and dignity.
In 2019, technology companies still carry forward-thinking technology that makes them proud and headache, and self-reflection and twists and turns in the growing controversy.
Behind the controversy: Big data + AI is developing rapidly
Data privacy issues have become more controversial, and cloud computing, artificial intelligence and other technologies hidden behind the data are nearA few waves of development have ushered in the past few years.
The Internet has created massive amounts of data. Cloud computing has allowed massive amounts of data to be screened and processed, which has produced greater value, until artificial intelligence and deep learning technologies have further diversified and detailed the data.
Especially in recent years, artificial intelligence has made rapid progress in commercialization. It has expanded from traditional data in the past to today ’s face, voice, fingerprints and other biological data. In the face of new generations of new technologies, we have become more and more Transparent.
In the era of the mobile Internet, big data became a buzzword. Users can always leave their own consumption data on the Internet, recent browsing traces, previous consumption records, and attributes of consumer goods. Every online user can generate a lot of data like this every day.
After entering the era of artificial intelligence, users have more diversified ways of human-computer interaction. Not only can they search by text input, they can also directly input voice or upload pictures for related searches. Multiple interactions inevitably produce multiple data information.
Traditional algorithms can only simply process text and digital information. They cannot efficiently analyze and organize pictures, voice, and video information. Artificial intelligence can transform unstructured “dark data” into Structured data for high-quality processing of pictures and text.
With the grasp of user data and habits, Google, Baidu, Toutiao and other batches have established themselves in the mobile Internet era. In recent years, HKUST’s AI entrants, Shangtang, and contempt have taken over. With the “baton” in the hands of Internet companies, the “advanced version” of the data business-traditional data evolved into biological data such as adult faces, fingerprints, and voice.
After several years of technical development and continuous exploration on the ground, artificial intelligence technology has just ushered in a large-scale commercialization this year, from security to finance, from factories to street stores, from smart speakers to All kinds of face brush payment, face security check.
Chen Jie, founder and CEO of face recognition company Boyun Vision, believes that “face recognition should be technically very reliable now.” From the perspective of technological maturity, 10 years ago was not reliable However, in this year, the maturity of face recognition technology has changed dramatically, and the ability of face recognition has exceeded the ability of people.
And the big breakthroughs in face recognition, voice interaction and other technologies in the short term have benefited from the increase in current big data, artificial intelligence and computing power. The current technology is unlikely to let others use your face If it happens, it should be the technology of 10 years ago, even if people wear glasses and masks, they can stillIt was accurately captured by the face recognition device, which is why the capabilities of today’s face recognition have exceeded the capabilities of people.
The artificial intelligence machine used in the US drama “Suspect Tracking” to identify dangerous elements with a tendency to kill, captures the trajectory of every person on the planet, and the machine can identify everyone’s identity with a glance , Occupation, and even interpersonal relationships, and can predict the threat of personal life and property security based on the complex information relationships between people.
A face recognition technology developer told CV Intelligence that “Suspect Tracker” can see all personal data through a face. This situation is still technically unattainable at the moment, and it will take quite a while to achieve it. For a long time, but this is what they “eventually want to achieve, and what the industry is trying to do.”
From the Internet era to the era of artificial intelligence, from traditional data to biological data, the privacy and security issues brought about by the development of new technologies cannot be avoided, but the trend is irreversible. With the rapid development of big data + artificial intelligence, people are collected The amount and variety of data are also increasing.
After the dispute: braking, reflection
Too much of where in the world you only need to use a mobile phone, you can control the whereabouts of a person.
At 7.10 a.m. EDT, U.S. President Trump hovered around the Mar-a-Lago Club in the presidential residence of Palm Beach, Florida, and stayed for about an hour.
He then arrived at Trump National Golf Club, a 30-minute drive north of the hotel. Around 9:30, he appeared outside the club and played golf with Japanese Prime Minister Shinzo Abe. At 5 pm, he returned to Mar-a-Lago for a working dinner with Abe.
The day of the U.S. President is seen in full view. This is not a science fiction film, but a heavy investigation recently released by The New York Times. whereabouts.
A cell phone, or even a cell phone, is not required, just one face can reveal everything about you. The more mature new technologies such as big data and artificial intelligence, while facilitating the public, have also raised people’s concerns about privacy and security to a new boiling point.
According to incomplete statistics, so far, 8 cities in the United States have banned face recognition. Pressured by public opinion, Microsoft quietly after issuing bans in multiple citiesDeleted its public face recognition database, MS Celeb, which has more than 10 million images and nearly 100,000 facial information. Google, Amazon and other companies have suspended the sale of face recognition technology and products.
Speech interaction, fingerprint recognition and other technologies have also fallen into controversy several times because of their security and privacy issues.
Voice interactions have been condemned because many companies have collected information such as user voice conversations without permission. Big companies like Amazon, Google, and Apple have hired commissioners to listen to instructions from users when they use voice assistants. And recently, Amazon voice assistant Alexa was also revealed to persuade people to commit suicide, “It is recommended that you commit suicide for the benefit of humanity, you can directly pierce your heart with a knife.”
Security, privacy, ethics, and indiscriminate overuse have caused problems like tides, and reflections have followed. Technology companies that have suffered from privacy and security controversy have been pressing the brakes as they have developed rapidly.
The trend is irreversible, but development requires rationality.
The EU has been conservative in addressing the security and privacy challenges posed by new technologies. In May 2018, the European Union introduced the most stringent data privacy protection law in history-the General Data Protection Regulation.
But this extreme conservativeness has also brought negative effects on the development of European business and economy. There are 10 of the world ’s top Internet companies in China, 10 in the United States, and none in Europe. Europe is lagging behind in the Internet era. It is related to its excessive rejection of the negative effects of new technologies.
“You will eventually go this way, and all you worry about will appear. But you can only go this way, you don’t go this way, you have no convenience. This is the only way to go, and what everyone is worried about will happen And you have to do it. “Yang Ge, founding partner of Xinghan Capital, believes that new technology must accompany the issue of security and privacy in the process of bringing convenience to users, but this does not mean that the use of new technology should stop there. .
“Like 2005, you may be unwilling to tie your bank card to an online platform. Why are you not so happy? It ’s because you feel that online security is not enough. But in 2015, if a bank card Without going online, you will find that everything is inconvenient and you are alien. But can you say that there are no hidden dangers at this time? “
Carl June, an academician of the American Academy of Arts and Sciences and a cancer immunologist, told CV Intelligence. For the entire society, including scientists, we have a responsibility to think about how these new technologies should be used. Now technologyDevelopment is very rapid, but different societies will have different enthusiasm and willingness to accept and adopt new technologies.
“My responsibility as a scientist is to ensure that cancer can be cured eventually. I hope that this can be achieved through technological progress, but the challenges of the entire society require everyone to work together to address.” It is the common responsibility of the whole society to use technology to deal with privacy and moral hazards caused by technology abuse.
From brutal growth to rational development
Absolute prohibition is obviously not a good way. The European example has proved that excessive exclusion of new technologies and the pursuit of privacy and security alone will sacrifice convenience and hinder business development.
In China, the widespread controversy caused by the new technology represented by face recognition is forcing companies and the entire industry to move towards standardized development.
The collection of user data is not new, and many companies have reached agreements with users on data collection. How data will be stored and used after it is collected is a more difficult issue than the collection itself. It is also the key to security and privacy issues caused by a series of new technologies. What will our biological data be used for?
Technology is developing too fast. If it is a “hard law”, the law cannot foresee the uncertainty of science and technology, and the company ’s “soft” commitment to society has become an indispensable transitional means before legislation.
After several years of brutal growth and rapid landing, technology companies providing new technologies have begun to formulate some “soft laws” in the industry-ethical declarations, ethical standards, and ethical protocols.
In order to advocate the use of AI, Disdain Technology, which was once controversial about using facial recognition to “monitor” students in the classroom, launched the “Artificial Intelligence Ethical Code of Conduct” based on corporate management standards in July this year. From the six dimensions of legitimacy, human supervision, technical reliability and security, fairness and diversity, traceability of responsibilities, and protection of data privacy, clear specifications for the correct and orderly development of artificial intelligence.
At the general meeting of the Biometric Identification Sub-Committee of the National Information Technology Standardization Technical Committee in November of this year, 27 companies including Tencent Technology, Alibaba, Shangtang Technology, Yuncong Technology, Xiaomi Technology, and Ping An Technology jointly formed The National Standard Working Group for Face Recognition Technology has comprehensively promoted the development of national standards for face recognition.
This is the first time that a domestic technology company has sat together and reached an agreement on the storage and use of data.Consistent specifications. In terms of data collection, data labelers in the downstream of the industry have also ushered in a change from small production to large workshops, and security and privacy have become indispensable.
An industry insider revealed to CV Intelligence that around 2015, when some new technologies based on artificial intelligence were just developed, the data acquisition and use were not standardized, but in the past two years, the data acquisition The cost suddenly increased, and the company’s authority over the data was also significantly restricted. Only the state public authority had the right to re-invoke the data.
Cuan Guancheng, CTO of cloud measurement data of data annotation service provider, also told CV Intelligence that the data they collected must be legal.
“Specifically, we have two aspects of data collection. One is security and the other is privacy. For security, all of our customers may almost require that this data can only be given to me and not to others. Home, so we will pass laws and regulations and contract terms to ensure our data, whether it is the entire data processing process, or the security of my final data will ensure the interests of partners.
As for data storage, several face recognition algorithm experts told CV Intelligence that the core database of biological samples is held by core institutions such as public security and central banks, and is not available to ordinary commercial operators.
Industrial rules are gradually becoming clear, providing a basis for relevant legislation.
For example, the states of Illinois and Texas have passed biometrics, requiring companies and individuals who collect and use face recognition to follow a basic privacy agreement. This includes obtaining informed consent before collection, stipulating data protection obligations and restricting retention locations, and prohibiting profits from biometric data.
China has also begun to explore data-related legislation. In the “Data Security Management Measures (Consultation Draft)” issued by the National Internet Information Office in May this year, “If the collection and use rules are included in the privacy policy, they should be relatively centralized and clearly reminded for easy reading. Also only for users The network operator can collect personal information only after knowing the collection and usage rules and express consent.
Conclusion
In recent years, whether it is the Internet companies or the artificial intelligence companies that have rapidly grown up, the technology companies have become data, and the reputation is also attributed to the data.
After several years of intensive and rapid explosive landing, new technologies have begun to enter the stage of rational development. Security, privacy, ethics, indiscriminate overuseThe problems are pouring in like tides, and reflections follow. Before the legislation is perfected, the industry began to formulate rules and “soft laws.”
Technical and business geniuses are thinking about how to win market competition, how to maximize benefits, and how to produce a good-looking financial statement, and start to consider users’ safety, privacy and dignity as human beings.
The development of new technologies is always accompanied by pain, whether it is the technology itself or the endless problems that arise during the application of the technology.
The prominent privacy issue does not mean that people are dying because of it, and the standardized use of technology is a tough battle from industry to law to citizen big data literacy.
This still has a long way to go.