Tale of AI in SEA

A Tale of ChatGPT and the Art of Negotiation

Picture this: you're locked in a tense negotiation with a vendor on a 6-figure deal, fighting tooth and nail to shave off the last dollar. You are up against a tough negotiator who knows that you need the deal, and yet you also have clear instructions from Finance about your budgetary constraints, of which this contract is threatening to explode. That's when you remember ChatGPT, the powerful language model developed by OpenAI.

Expectantly, you turn to your laptop and start typing away. Minutes later, you're presenting your case with the confidence of a seasoned negotiator, backed by ChatGPT's deep understanding of the market and its mastery of the art of persuasion. To your amazement, you secure a deal that is 60% better than your initial offer.

That negotiator was me, just a mere 2 months ago. As I walked away from the table, I couldn't help but wonder: am I, as an entrepreneur, now automated away?

The Harsh Reality of Layoffs in Southeast Asia

The truth is, Southeast Asian companies have not been immune to the recent economic slowdown, and the waves of layoffs have already hit the region. Many companies from the publicly-listed unicorns to decades-old SME had been tightening their workforce investments. With that belt-tightening also comes natural questions about who will get the job done. Aside from the age-old solution of contracting and offshoring, the somewhat unfortunate timing of ChatGPT crashing into the scene also had many business owners thinking hard about its potential as replacement for certain classes of knowledge work. On the other side of the table, these rational questions take the form of simmering worries by employees about what it means for their profession.

The Fear of AI Taking Away Jobs

This fear is not new. Throughout history, new technologies have always been met with resistance and concerns about job loss. For example, the Luddites of the 19th century smashed looms in protest of the new machines that threatened their livelihoods. Today, many pundits worry that the rise of AI will lead to widespread job loss, just as the rise of automation did in the past. On the ground, at least among people I know, that fear is a lot more muted, coming more in the form of a "wait and see" curiosity. That is because for all the pomp and circumstance of large language models, it is not the latest technological shift that was heralded with world-shaking potential, only to land with the same incremental impact that characterise the majority of real-world adoption.

Even in the realm of AI specifically, it is not the first time it has been applied in actual business context. In the 1980s and 1990s, AI gained traction in the financial services sector, where it was used for fraud detection, risk management, and portfolio optimization. And yet, even in those sectors, we still retained our fraud analysts, risk managers and portfolio managers, albeit with a requirement for more savvy in leveraging these new tools.

The Reality of AI Adoption in Southeast Asia

When it comes to AI adoption in Southeast Asia, the picture is even more nuanced. Despite the decade-old hype around big data and machine learning, the penetration of these technologies is still relatively low in the region. In many cases, the limiting factor for AI adoption is not the state of the technology, but rather the maturity of the organisations here.

As a tech entrepreneur myself, and having spoken to many fellow startup founders, I have seen this limiter play out firsthand. Unless machine learning is a core differentiator, and it is in a lot less cases than what our pitch decks suggest, the background investment in a robust data infrastructure and a high-standards data management culture and stewardship is usually kicked down the priority list behind more existential concerns of survival and growth. Anecdotally, I have observed that this is the norm among the SMEs that dominate our SEA economy.

Embracing the Future of AI and Work in Southeast Asia

If recent history of technology adoption is any indicator, the more pertinent question is not how AI will take over our jobs, but rather how we might under-leverage it, which is the likelier scenario. Behind the deployment of a flashy machine learning-powered product, there is a looming shadow of architecture, operations and culture that supports it. Similarly, for large language models to be effectively employed in our day-to-day operations, we have to recognise that it is not plug-and-play- there is a level of organisational understanding that goes into maximising its potential.

As for my little vignette on the ChatGPT-powered negotiation, I have to come clean that I was still the one driving the objectives and boundaries of my arguments, and ChatGPT helpfully, but also in a sense, merely fleshed that out into coherent paragraphs. Instead of spending 30 minutes tweaking phraseology, I pumped out a draft in 10 instead. So I guess for the foreseeable future, I will still be keeping my job, or at least 1/3 of it.