Cloud Hosting

Artificial intelligence: its one of those mind-bending subjects that provokes fascination yet is misunderstood by almost everyone (myself included). If you bring it up in the pub to sound brainy, the conversation might only go as far as Alan Turing, Deep Blue versus Garry Kasparov and perhaps the new series of BBC drama The Capture, in which foreign spies seem to use AI to manipulate CCTV and live broadcasts.

But AI has long moved beyond the realms of gripping fiction and headline-grabbing experiments into the real world. For instance, DeepMind, a British subsidiary of Alphabet, recently released almost the entire make-up of the protein universe, mapped for the first time by its AI programme AlphaFold. This data is already being used to advance our understanding of life-threatening illnesses and help protect honeybees.

Amid all the scaremongering about AI, perhaps we should be more aware of these positive developments, particularly if were investors. Funding technology that is tangibly solving the worlds trickiest problems surely thats responsible investing at its finest?

Besides, AI applications that drive efficiency, reduce costs, serve vital needs, and cement a companys competitive advantage should generate significant value. Thats not to be sniffed at in this current investing climate.

The Sanlam Artificial Intelligence Fund has specifically hit on AI as a way to back winners in the stock market. Its managers Chris Ford and Tim Day marked the funds fifth anniversary recently by sharing some intriguing insights on how AI has improved in recent times.

Things that people called pipe dreams when we launched the fund in 2017 are now happening, says Chris Ford. What was impossible is now possible.

Examples include Alibabas equivalent scoring better than humans in Stamford Universitys reading and comprehension test, and the Canadian platform Blue Dot spotting the spread of Covid-19 nine days before the World Health Organisation put the alert out.

How has this translated into hard financial returns? The 715.5m fund has had an annualised return of 18.2 per cent since it was established, which is respectable, and its top performers have included household names like Tesla, Netflix, Microsoft and Ocado, alongside lesser-known AI innovators like NVIDIA, Zendesk, and Appen.

The managers even have their own AI application (Orbit) to help them stay across things. Its ongoing charges figure is 0.8 per cent not dirt cheap, but not bad for an active fund.

Do actively managed thematic funds like this have the edge, not just over their cheaper rivals but within the broader investment universe?

Sanlams fund, a traditional open-ended unit trust, has comfortably beaten the MSCI World index, which has had an annualised return of 8.41 per cent over the past five years. By contrast, only 39 per cent of thematic exchange-traded funds (ETFs) have survived and outperformed the index over that same period, according to data from Morningstar.

But the funds performance has markedly dipped in recent times, as the global rout in stock markets hits technology shares particularly hard. This is the downside of thematic investing: clever use of AI may not be the main reason why a company flourishes, or why investors feel well-disposed towards it.

The likes of Tesla, Ocado and Netflix can end up in choppy waters when investing conditions change, and it becomes clear that these firms have been overvalued by the market for what they do. Thematic investing can also lead to a good deal of overlap in your portfolio, with the same names cropping up again and again. This can undermine diversification and add to your risks.

Also, whilst many of the Sanlam holdings are doing great work (like Alphabets DeepMind), there is a good reason why its not classed as socially responsible. EthicsGrade is a ratings agency specialising in understanding the risks stemming from digitalisation, particularly AI, and it has awarded a top A rating to only a handful of firms, including Microsoft and Deutsche Telekom.

Fewer than a third of the 302 major companies it graded have a score of C or above. Too many were unrated altogether because they either ignored or provided little to no details on governance and ethics policies relating to AI.

Experts mostly agree that the more sinister AI possibilities as depicted in The Capture still look far-fetched. But its worrying that many of the major companies we buy from and invest in are being cavalier regarding privacy breaches, data misuse, racial and gender discrimination, the creation of harmful weapons and other potential dangers of AI.

AI is not going anywhere (particularly since the Chinese government is investing heavily in AI research and development). If it proves as beneficial for its first movers as many believe, AI could become the first big success story of so-called thematic investing, previously dismissed as no more than a trendy gimmick that promises more than it can deliver.

The big question is whether this technology will be treated with enough care by its creators and investors. If not, disappointing returns will be the least of our problems.

Originally posted here:
Investing in artificial intelligence: The pros and cons - iNews