Deep Dive Podcast: DeepSeek - What difference will cheaper AI make?
DeepSeek has shown that AI innovation is not limited to big tech players with billion-dollar budgets, and as models get better, expect to see more flux, say industry experts.
CNA's weekly news podcast takes a deep dive into issues that people talk about at dining tables and along the office corridors. Hosted by Steven Chia and Crispina Robert.
This audio is generated by an AI tool.
An AI model developed by a little-known Chinese hedge fund that reportedly rivals OpenAI's ChatGPT at a significantly lower cost has triggered a deep “freak out” among the global tech industry. Why was there a big reaction to DeepSeek and what are the implications for Singapore?
Dr Leslie Teo, senior director of AI products at AI Singapore and Jianggan Li, founder and CEO of venture outfit Momentum Works break it down for Steven Chia and Crispina Robert.
Here is an excerpt of the conversation:
Crispina Robert, host:
We can't talk about this without talking about how it will impact us, right? In terms of jobs, professionally, for example, we're all deep into this. Every company I know is experimenting: How is it making your work more efficient? How can you do things better?
And some people are saying it's even displacing some very junior-level jobs which ChatGPT can do. How worried are you on a scale of one to 10 (about the potential impact of AI on jobs)?
Jianggan Li, founder of Momentum Works:
Two.
Dr Leslie Teo, senior director at AI Singapore:
I'm at 10.
Steven Chia, host:
Okay, now tell us why?
Jianggan:
I think Leslie kind of answered my question, right? Because if you have people like Leslie who worry at the level of 10, then I don't have to worry.
Crispina:
(But) you're optimistic that humans will adapt?
Jianggan:
I'm optimistic for the long term. I'm a little bit cautious about the short term, because I (did) see that when ChatGPT came out, when Midjourney came out, I (did) see (that) ... gaming studios started laying off people because graphic design and certain functions, they could do it more efficiently, and they (needed) fewer people. And we can always say that, okay, the skills can be retained.
I mean, you open new opportunities, but there's always a gap between the time that jobs get displaced versus the time that new jobs get created, and it's painful for people. Especially people who have been trained as (specialists).
Steven:
But you're still confident that, overall, humans will prevail.
Jianggan:
I'm kind of a technology optimist. I look at global warming. I don't like it, but I'm not terribly worried about it. I'm hoping that someone (will) find the technology to solve it.
Leslie:
Technology is generally a positive thing. Economists are really positive about it, but this speed is very fast, so it's a timing problem.
For me, it's good to be at 10, because I want to make sure the technology is used for human good. Policymakers will say AI is to augment not replace.
Unfortunately, the profit motive is: "What are you talking about?" Let's be truthful about this – it's to replace. And this is happening so fast, right? I am worried we can't absorb this actually. So it's better to be more alarmist in that situation.
Crispina:
So in other words ... we'll be forced to figure out how to protect ... the transition (to using AI more in workplaces) right?
Leslie:
I think the transition is important.
People say the Industrial Revolution was a good thing. Yes, it was, but two generations of workers suffered before the good thing happened.
