Asia’s AI transparency tangle

Sign up for the Daily Newsletter, sent exclusively to our premium subscribers. We break down the big and messy topics of Asia’s tech and startup community. Get the newsletter in your inbox everyday with a premium subscription.

Hello reader,

I was a pretty terrible English teacher in my younger days for various reasons, including an inability to always explain the rules of English grammar.

It’s an issue a lot of native speakers face, knowing what correct grammar is but being unable to explain the rules that make a particular sentence structure correct.

Today’s increasingly powerful AI systems are similar in that even their creators struggle to explain how these models arrive at the outputs they do.

But while I could simply tell my poor students that I was right just because (I told you, a terrible teacher), it’s an important ethical issue for those impacted by AI systems to be able to understand them. Today’s featured story dives into that dilemma, with a focus on how Asian nations are approaching it.

Today we look at:

• Dive into the AI transparency paradox in Asia
• South Korea’s Kakao enjoys a bumper Q2
• Other newsy highlights such as Pinduoduo founder Colin Huang being named China’s richest person and Singapore-based UnaBiz getting US$25 million in pre-series C funding.

Premium summary

To be or not to be transparent?

Image credit: Timmy Loen

Building transparent AI models is difficult and costly, and it may also compromise accuracy. But it’s a key ethical issue in AI, a sector that has a huge impact on all facets of life.

However, the way Asian societies are approaching AI development could end up compromising transparency in favor of national interests.

• Paradox: In general, AI transparency means making the decision-making processes of AI models accessible and comprehensible to users, developers, and other stakeholders. However the complexity of many AI models can sometimes mean that even their creators cannot fully explain the decision-making process, and illuminating it can be overwhelmingly expensive.
• Conflicting approaches: Most countries around the world agree with the need to regulate AI and its far-reaching consequences, but they differ on how to put rules in place. The European Union, for example, has laid down transparency as a general principle, so users must be made aware of when they are interacting with AI. In contrast, China has unveiled an AI platform infused with President Xi Jinping’s thoughts and other themes championed by the ruling party.
• Sci-fi come to life: Defining regulations for AI systems is important now, but if sentient AI is ever created, it could become an existential issue. While that possibility may seem remote, some countries are already grappling with what rights such machines should be given.

Read more: National interests trump AI transparency in Asia

Kakao ka-ching

An honest look at how genAI translates into tangible results

Quick bytes