I wish I could press AI’s ‘delete’ button – Wierzycka

You can also listen to this podcast on iono.fm here.

This interview was originally aired on RSG Geldsake with an Afrikaans intro, and has been transcribed below.

RYK VAN NIEKERK: Magda Wierzycka is the founder and CEO of Sygnia, and she is known for not mincing her words. She often looks at major market and social developments from different angles that are both contrarian and commercially relevant.

This was again the case during Sygnia’s latest results presentation yesterday. The company reported a 25% rise in pre-tax profit of R260 million. It was for the six months to the end of March.

Read: Sygnia lifts interim dividend as AuM top R460bn

But it was Magda’s comments on artificial intelligence that stood out. She warned that AI would significantly disrupt the world we live in, that it could improve productivity, but that it will lead to major job losses, undermine human privacy and increase the risk of cybercrime.

Watch/read:

At the same time, Sygnia is not standing back from AI. Magda confirmed that the company is already aggressively rolling out AI solutions in the business, throughout Sygnia.

Sygnia has also launched the Attentia Fund, a new AI fund which will invest in AI infrastructure and related technologies. Magda is on the line.

Magda, thank you so much for your time. There seem to be several contradicting statements. First, you are warning that AI will change the way we live – and not always in a good way. Well, you are also investing in AI, and you want to persuade your clients to invest in AI.

MAGDA WIERZYCKA: Ryk, I’m always a world of contradictions. AI is already disrupting the world. It’s already disrupting the job market, so it’s current tense, rather than future tense.

The reality of it is that the genie is out of the bottle.

If I could press a giant delete button on AI as an innovation, I would, because as much as you can talk about health breakthroughs, scientific breakthroughs, you can talk about productivity increases.

You are also talking about a sudden massive concentration of wealth and power in the hands of the few, and mass dislodgement of entry-level jobs – and in fact white-collar jobs. Unfortunately, it is what it is.

So yes, we are embracing AI within Sygnia. But, as I said, and what I’ve promised all my staff, is that not a single person is going to lose their job.

But I see Sygnia transforming where the technology side is an enabler and our staff just become hands-on ambassadors for all our clients. So that’s the one side.

And then on the investment side, again I like to think that we are leading-edge innovators in product, so we have launched a Sygnia Attentia Fund. It is a little bit because of capacity constraints.

We’ve been running it very quietly for the last couple of months and, when I’m able to, I will talk about it more widely. There are big plans for the Sygnia Attentia Fund, and we already have investors and committed investors. But I’m not ready yet. Give me a month and I will be able to talk about that fact more publicly.

But in terms of AI, I just one comment. Someone said to me, well, if I believe – and that’s the world of contradiction – that AI is so bad for humanity, why am I investing in it? My answer is: ‘If responsible investors don’t invest, it’s a little bit like ESG.’

If responsible investors who can rein in AI deployment – and not the investors in those companies – then we have a completely unregulated Wild West.

RYK VAN NIEKERK: Yes, I hear what you’re saying, but disruptive innovation is not new. We’ve seen how the printing press and the steam engine and the internet changed the world we live in and work in – and in those examples the world was better off afterwards. Is AI different?

MAGDA WIERZYCKA: Because it’s not only touching on mass displacement of jobs, it’s also touching on numerous other risks that I didn’t list in my statement and comments.

Let’s talk about children. Children are at risk to be blunt, uneducated, and stupid, because if you can go home and your homework is being done by ChatGPT – including science and maths – then where is learning?

Where is innovation? Where is creativity? So your dependence on the world of AI is growing – and that’s cognitive development.

Then you’ve got issues, such as why form social connections if you can go home and you’ve got this delusionary, imaginary 24/7 available best friend in ChatGPT that always tells you what you want to hear.

So you are bullied at school, you come home, you have ChatGPT – which reinforces everything that you’ve done.

So now what you are seeing in children is a decline in empathy levels, because your social connections are disappearing and are reducing. That is something that behavioural therapists are already talking about, and what I heard at Davos.

So there are other indirect consequences, not just job displacement.

RYK VAN NIEKERK: But is it just generative AI, because ChatGPT was launched in November 2022, not even four years ago, and look at the impact it has had. Just imagine what’s going to happen over the next decade or so. Do you think this will accelerate?

MAGDA WIERZYCKA: 100%. Let’s look at what has happened in the last four years. It is unfathomable. When ChatGPT was first launched, it really was an enhanced version of Google, whereas now we are looking at just one AI.

But we’ve gone from basic conversations to advanced analysis and planning, from text-only interactions to images, audios, documents from very poor mathematicians to brilliant mathematicians, mathematicians and analysts who can talk in over 100 languages.

We are no longer talking about just text. We’re talking about voice. Where do we go from here?

If this is what has happened in four years, what happens in the next four years? We already have these AI agents or digital workers making autonomous decisions which are not controlled by humans. They haven’t yet touched the physical world.

But let’s look at Argentina and the latest reform, where you are now allowed to register a company as an AI agent [chuckling], and your board of directors is just AI agents, not people. So what happens next?

RYK VAN NIEKERK: Yes, it’s going to be an exciting journey.

I just want to return to the investment part of AI. We are going to see the listings of SpaceX, Anthropic and OpenAI very soon. They are going to be absolutely massive, massive listings. What do you think the impact of these listings would be on the AI investment world?

MAGDA WIERZYCKA: Well, I think it’s twofold. Obviously these listings are going to attract a lot of money. But these companies are capital-hungry, because what’s driving all the deployment of AI is massive energy consumption.

So these companies – as much as I mean Anthropic has multiplier growth in their revenues – so does OpenAI. But they have to continue spending that money on data centres and AI infrastructure.

Interestingly, there’s one good development that comes out of that, and that is the focus on renewable energy, because something needs to power these data centres. So there are secondary effects. Other companies, listed companies, are already benefitting and will benefit in future. And new companies will be formed as a consequence.

But these companies undoubtedly will attract – and have already attracted – enormous amounts of capital and that capital will be diverted away from other investments as, again, we are already seeing. That’s why I said it is current tense.

If you look at Anthropic, a year ago it was valued at $61 billion. A year later its last valuation is $950 billion. OpenAI, still in the private space, understood, $300 billion a year ago and $850 billion the latest valuation. These are insane numbers.

RYKA VAN NIEKERK: Are they realistic valuations, though?

MAGDA WIERZYCKA: Yes, I completely believe, if you look at what these numbers are driven by, they are not driven by hysteria of listed markets. This is all happening in the private markets, where investors are not a retail investor who has that ‘fear of missing out’.

People who are investing in OpenAI set these valuations – Anthropic, Amazon, Microsoft, Nvidia.

It’s a little bit of a circle of the same players playing the same game and investing in each other, and at the same time buying services from each other. So this is where you can see the power concentration happening.

And we are only talking about the West. Heaven knows what’s happening in China. You’ve got an idea that the same level of innovation is happening in China. It’s just less transparent.

So yes, I think the moment those companies hit the listed markets those valuations are going way up.

RYKA VAN NIEKERK: Do you think the investment opportunities are only in those companies because, yes, they develop platforms and infrastructure for the development of AI and applications thereof?

But there are also many companies in the financial sector, in the medical sector, research sectors that will benefit significantly from proper AI deployment. Do they not also offer investment opportunities under the AI umbrella?

MAGDA WIERZYCKA: Well, that’s an interesting question because, again, AI-type analysis is already being used in data crunching; it is already being used behind the scenes in, for instance, many medical devices.

But the more that happens, the lower the cost should be. What is medicine pricing based on? The amount of money you have to spend on research, and the years that it takes to research a new drug.

Now, if you can speed it up – and let’s talk about quantum computing, which will be just AI on steroids – then the cost of that medication should come down. If the cost of medication comes down, the profitability of these companies should come down. So, it’s a bit of a double-edged sword and it still is a bit of a great unknown.

But one would expect with this level of innovation and, we are already seeing it in software development world. At Sygnia, we’ve been in software development for the last 20 years. I’m just observing how much faster we can develop software.

And of course, the next question from clients is, ‘Where is the fee reduction?’ Thank goodness we are already a low-cost provider because margins hopefully are not going to be under as much pressure as others. But what are we talking about? Lower profits.

RYKA VAN NIEKERK: Is AI managing your funds, or do you still have a warm body behind the final decision making?

MAGDA WIERZYCKA: Well, a lot of our funds are index-tracking funds, so theoretically we do have warm bodies; but they work in a very quantitative space.

So the AI in the investment world [makes] a lot of our index tracking funds obviously less relevant, other than the speed of implementation of decisions, the speed of analysis. So that works.

Then we do run multi-asset classes, funds like our skeleton funds, where we do invest in various asset classes in one bundled portfolio. Although each asset class is represented by an index, ultimately the mix is determined by human beings.

Listen/read:

At the moment it is very much warm bodies making decisions, but supplemented by research that they can now do and the access to information that they now have through things, provided you pay the highest subscription levels through, for instance, ChatGPT and the Claudes of this world.

RYK VAN NIEKERK: Just lastly, I want to come back to the Attentia Fund. You said you will speak to us about that a bit later. Just the name – Attentia? Does that come from the Latin for attention, because if it does it seems to be a very interesting name.

MAGDA WIERZYCKA: It takes a bit of my innovation because – if I can brag a little – it kind of talks to the future of Sygnia Attentia Fund, which I won’t talk about now.

But the name came from the very first paper written on AI, which was written by kind of eight researchers who invented this concept of a transformer – which really is the architecture behind large-language models – and that paper was titled ‘Attention is all you need’. And that’s the word Attentia.

I’ve now moved in terms of the Sygnia Attentia Fund, and its kind of slogan is from ‘attention to intelligence’.

RYKA VAN NIEKERK: Well, my high school Latin came in there.

But anyway, let’s leave it there. Magda, thank you so much for your time. That was Magda Wierzycka, the CEO and founder of Sygnia.