Credit: Mohammad Hassan/Pixabay

The waves of technological innovation have intensified, yet some still stand on the opposing side of change.

SY Lau, Senior Executive Vice President of Tencent retold a story to TMTPOST. Back in 2009, a visually impaired user was chatting on QQ, the live communication tool that made Tencent a name in the industry. During the chat, the person with whom he was chatting sent an emoji. Since the screen-reading software at the time could only render emojis as blank spaces, the visually impaired man felt greatly troubled, believing he was being mocked by the other user.

After later receiving the feedback from the user, the QQ team got to work to improve the emoji reading function on their mobile version in order to allow blind users to identify expressions sent to them while at the same time convey their own thoughts and feelings using expressions like anyone else.

Henceforth, QQ's screen reading functionality was not only limited to reading emojis. Harnessing Tencent's AI capabilities, QQ now enables visually impaired users to easily 'read' text sent by their friends.

"Perhaps you don't consider what you have done as remarkable, but what you have done has made it possible for me to hear and feel," came the feedback from one user.

The priority challenge for AI: smart tech or human benevolence?

The development of new technologies, the internet and digitization is deepening. Social organizations, structures and operations at the macro level, together with human lifestyles, values and ideals at the micro level, are undergoing subtle but profound changes.

No doubt AI has already penetrated deeply into the daily work routines and lives of human beings. Yet at the same time, it attracts a considerable amount of fear, criticism and doubt.

The tide of technological optimism is now, inevitably, receding, as people thankfully return to the realization that the requirement of human benevolence should always outweigh the importance of intelligent tech. No doubt AI is a powerful tool and aide; however, it must be developed responsibly and with human guidance, in both caring and humane ways.

At the AI Everything Summit held last week in Dubai, SY Lau, in his role as one of the representatives of technology companies, proposed the concept of "responsible AI".

SY Lau, Senior Executive Vice President of Tencent and Chairman of Group Marketing and Global Branding, speaking at the AI Everything Summit in Dubai

"To minimize any of AI's adverse ramifications while on the path to sustainable development, we need not seek out neutral technology, but instead technology for good and what I call 'responsible AI'", stated SY Lau during his speech.

The road from 'respectability' to "tech for good"

This was not the first time that Tencent conveyed the concept of "tech for good". There are instances of its reference being made in both internal and external speeches of company senior executives.

In a speech delivered at the 2018 Tencent employee annual conference held late last year, Allen Zhang, President of WeChat Group at Tencent, said, "Benevolence is more important than intelligence. AI is smarter than you. However, you can be a lot kinder than it can."

Only several days ago, on 6th May, at the Building a Digital China Summit in the city of Fuzhou, Pony Ma, CEO of Tencent, officially named "tech for good" as the future vision and mission of Tencent.

In an interview to media outlets that included TMTPOST, SY Lau mentioned that the "responsible AI" he advocated for and the concept of "tech for good" proposed by Pony Ma coincided, even though they did not exchange ideas beforehand. This has now become an important topic for the top-level management at Tencent who are now writing the concept of "tech for good" into the company's charter of values.

Lau recounted to TMTPOST that roughly two weeks ago, the team of top management at Tencent scheduled a meeting to discuss some ideological issues at the company such as "What is the soul of Tencent?" and "What are the company's visions and mission?"

"I have worked for Tencent for 13 years, devoting most of my life's work to the company. However, in these 13 years, people's attention has been focused on business and product-related issues, with very little regard paid to ideological concerns," SY Lau related to TMTPOST.

Lau recalled that during the meeting one person posed the question of whether the company should boldly embed "tech for good" into the vision statement of Tencent.

The top management team went on to debate this question for nearly four hours. The key concern was whether it was too early to propose such a value, or if that such a value was indeed achievable at all.

Later, Pony Ma stood up to tell everyone that at the time he founded Tencent he desired to build the "most respected" internet company. "In the process of building the 'most respected' internet firm,” he said, "we also received a lot of criticism, but now is definitely the right time to pursue the mission of 'tech for good'. If we, the people and product managers at Tencent do not believe in the concept of 'tech for good', we are in big trouble."

SY Lau believes that this discussion was truly necessary since it advocates a direction for newcomers to the company while offering an important heritage that the founders should pass on to the rest of the firm. Future employees of Tencent should be made aware of what he has dedicated his life's work to.

However, this raises the question of what criteria should be applied to judge what is good or evil in the context of a complex system in which technological upgrades are intertwined with commercial interests.

In the past, Google proposed the corporate principle of "Do the right thing: don't be evil", even going so far as to include it in the prospectus of its IPO back in 2004.

Google claims that the principle of "don't be evil" represents one of the core pillars of its corporate identity and a cornerstone value of the firm. For a time, it assumed the form of a "code of ethics" that was highly valued in the world of tech.

However, within Google, the principle of "don't be evil" has languished. In 2015, Google's activities moved into the war arena when it participated in Project Maven, an AI project of the U.S. Department of Defense utilized to improve the effectiveness of drone attacks in battles.

Although the opponents of the final Maven plan finally won over, the words "don't be evil" have been removed from the beginning of the Google Code of Conduct, with only one reference retained towards the document's end.

Then what are the AI moral standards in the eyes of Tencent, and what principles should be adhered to in keeping with the vision of "tech for good"?

SY Lau says that the "tech for good" now being advocated in an Eastern country differs slightly in essence with that of "don't be evil" proposed in the West, that the concept of 'good' should not be stipulated by laws but rather by cultural consensus. According to Lau, the concept that men are born into this world 'good', an idea propagated since ancient times, is an idea worth advocating in the era of AI. What tech companies need to do is to use technologies to maximize the goodness in people; it should not only be about refraining from doing evil, but instead advancing towards a better and just world for all.

For instance, many tech companies use technologies for middle-tier development, and data to mine business value. However, AI should assist in making information more transparent and big data should play a greater role.

In Tencent, where AI has already penetrated the underlying technologies, the basic principles to adhere to can be summarized by the acronym ARCC: Available, Reliable, Comprehensible and Controllable.

More specifically, in terms of application, and taking SY Lau's forte of advertising as an example, AI should be able to save people time in their purchase decisions that will help them better manage their household finances, instead of wasting time on useless items or items whose cost price far exceed their value. "In my mind, this is the advertising utopia."

 Credit: Daniel Stori/{turnoff.us}

Counterbalancing regulation and innovation in AI

However, another perspective also exists that with AI still in its early phase, any over- discussion concerning directions or principles will only serve to stifle advancement of the technologies that are being developed. Setting up rules of the game at such a premature stage may kill off technological innovation; instead we should let AI first run its course and prescribe any rules at a later stage.

In response to this concern, SY Lau told TMTPOST that the relationship between technological advancement and human ethics is not, in fact, a case of "this or that" but of "this and that". Sometimes formulating the rules first then standardizing the principles and directions in a structural and orderly manner will be more conducive to the healthy development of a field that has such enormous potential.

Advances in technology are going to spur the whole society on towards new modes of operation, modes of operation that are going to shake the traditional foundations of our human lives and lifestyles, from communications to entertainment, and even trade and travel.

Compared with previous technological revolutions, digitization, and particularly AI, may ultimately lead us to the creation of an information gap that will only further fragment the human population, exacerbating inequalities and posing huge challenges to society.

New technologies have replaced traditional tools. For those of us who are not endowed with new skills or discernibility, it means deprivation of power, the negative spillover that comes with technological progress. For this reason alone, tech companies need to act responsibly.

International agreements must exist, which may then evolve into general principles, then developed at a later stage into concrete regulations and laws. The European Union's General Data Protection Regulations (GDPR) is a good example, being the world's first legislative attempt to address the important issue of data security, prescribing data rights for individuals and imposing security responsibilities on the managers of our data. To some degree, it sets constraints on the tech industry. Even though right now it is still difficult to reach global consensus, the GDPR provides a set of indispensable guiding principles.

For Tencent, whose products are used by hundreds of millions of users, the ideas of "inclusion for all" and "accessibility for all" must be practiced and propagated under the premise of a 'fair' algorithm for all.

The above-mentioned story of QQ providing the solution for visually impaired users is just one example. Due to the huge population base, the number of disabled users in China is by no means a small minority. According to statistics published by the China Disabled Persons' Federation, as of the end of 2010, the total number of disabled people in China was 85.02 million. Dependent on each person's specific physical or psychological limitations, they will more or less show a relative lack of interest in new technologies compared to that of the general population.

However, accessibility features can lower the acceptance thresholds of new technologies for disabled users. In this age of seemingly overnight digital reform, no longer need they be forced to play the part of a missing party, but able to participate in the adoption of mainstream technologies and join in the waves of technological innovation such as AI that are sweeping the globe.

SY Lau added that "All this talk is not intended to publicize Tencent. Instead, it is a means to accumulate a theoretical approach that can be adopted in real-case scenarios and optimized by tech companies across the board."

-----------------------------------------------

This article was edited by @Dido Pang . The original Chinese version was written by Li Chengcheng . 

Follow us on Twitter @tmtpostenglish, Medium @TMTPOST and Facebook @TMTPOST.