There is a bubble in artificial technology stocks.
What I would be interested to what industries or companies people think will be the largest beneficiaries and conversely perhaps the biggest losers.
For example, Sonic Healthcare - the ability to leverage AI for blood tests.
@Mujo I think any company that has access to large amounts of data. So I included infotech firms that capture client data via their SaaS offerings but also firms that have made big investments in digitalisation as well as holding customer data.
Examples only:
Infotech SaaS AI winners include: $PME, $WTC, $XRO at the bigger end, and also $RUL, $CAT, $AIM at the smaller end
Big "digitalisers": $CBA, $BHP, $RIO, $REA, $CAR
Holders of large amounts of customer data: $WES, $SUL, $JBH
Engineering Tech firms with niche applications: e.g., $DUR and their MEND technology service is one example in my portfolio
So my thesis is that those firms that have been on the front foot of digital technology adoption will be best placed to leverage the data sets, including longitudinal data (time based) to apply AI and ML to drive further insights both to run their business better (more efficient; better capital allocatio) and/or to offer more value to their customers via efficiency and effectiveness.
Overall: the strong will get stronger.
I heard a recent Goldman Sachs podcast about the AI revolution (... or bubble, depending on your point of view). If I recall their analysis correctly it goes along the following lines:
Given the capital that's being deployed, and the lofty SP expectations, you may be right in that we have to pass through a valley of disillusionment, before we see success at scale. I don't know.
What I am observing is that the historically capital light tech leaders like $MSFT, $AAPL etc. are becoming much more capital intensive.
But the "application layer" like $PME, $WTC, $XRO,....$CAT, $RUL, even $AIM etc. are scaling in a capital-light fashion.
My bets are on this application layer,....but I'm still exploring, and so everything I've written above is more of a hypothesis, than a well-baked point of view.
Great topic for discussion.
I think i'm on the same page @mikebrisy -- but so hard to know at this very early stage. I wonder if there's anyone who is deep into this we could interview? If you can think of someone, i'll do my best to schedule a chat with them.
And I think you're right too @Mujo -- there's plenty of companies talking up the advantages of AI that really wont see any enduring advantage from the tech.
Sorry typed the question on my phone hence the horrible grammar.
The software and customer data def makes sense @mikebrisy thanks. The software companies do seem to be the major winner again.
I agree if everyone in the industry gets access there is no new competitive advantage, and where most will end up @Strawman Then again depending on the competition dynamics there could be a step change in the whole industries profitability - if they don't pass on margin to customers. Where one company has a larger amount of data they may be able to leverage AI more efficiently too I guess.
George Lee, co-Head of Goldman Sachs Global Institute, is deep into this. Alas, I don't bump into him very often, partly because our private jets don't cross paths that often. (LOL) (For the record - I don't know him at all)
Here's a clip to help frame the thoughts: https://www.youtube.com/watch?v=ZlEfAR11i1o
Many years ago, when I was responsible for corporate transformation at a large multinational, I was advised by a Boston University Professor by the name of John Henderson (now long retired). He'd done research on the IT revolution of the 1980s-1990s and found that technical innovation happens quickly, and there is an over-reaction of enthusiam to that. However, the big value "use cases" require technical innovations to be embedded in business processes and also for the behaviour of people to adapt. This takes years to unfold.
Of course, disruptive technologies can blow up business models (as we've seen in too many examples to even cite here). But even then, the disruptors generally have to build processes and organisations (people!) at scale to deliver material value.
In that respect, my hypothesis is that AI will be just like the first IT revolution, then the internet-revolution, and then the SaaS revolution.
Human behaviour and the time to scale business processes will create a drag on value delivery. The life cycle may be shorter than before, because increasingly (because of the previous IT revolutions) capability can be deployed at global scale with fewer people. But the adaptation by people will still be part of the realisation of value, and that's the slow part.
In other words, while capital can be deployed quickly, behavioural change is slower, and organisational behaviour change slower still.
And so, just as with IT-systems, the internet, and SaaS, there will be winners and losers, because people, behaviour, culture, leadership will yet again be differentiators.
Execution will vary hugely.
There is an argument that once everyone has a ladder, no-one has an advantage from being able to see over the wall. But I think that's missing the point. Some will get their ladders in place quickly, others will try and fall off them, and others will take longer. I don't know what the timeframe is for everyone to get their ladders in place. But because it involves leadership, people, culture, behaviour ... the difference between the first, the middle and the last, will be years, and perhaps many, many years. John Henderson's research contended that the timescale was actually decades!
Maybe what I've written is just so generic and obvious, that it is not insightful at all. But hey, it's getting on, on a Friday afternoon.
Thanks for sharing that clip, @mikebrisy, it really resonated for me. As did your post.
The adaptation of people really is the slow part, and that's hard to endure when you see it early.
The commercialization of electricity started in the 1880s, and although some were excited by its potential, it was not widely embraced at the time, being considered expensive, dangerous, and unreliable by most businesses. It took decades to gain real traction.
A similar story unfolded with radio, steam, cars, aircraft, and, as you say, more modern disruptive tech.
Even with widespread adoption, the value certainly didn't accrue evenly -- and often in unexpected ways.
The lesson? It takes time for disruptive tech to make its mark, even when it's already largely proven.
I'll recommend the book "The Gorilla Game" again here, which puts all this into an investing context.
Basically, wait for a clear leader to emerge, then invest heavily and stay with it for a long tiime.
@Strawman Nick Griffin from Monro partners has just returned from a business tour of the US. He has one of the best investment records in the tech space over the past 10-15 years. Alex Pollak from Loftus Peak is another fund manager I respect.
@Mujo probably not too much to add to Mike and SM. the most coherent forward looking strategies i have heard revolve around three stages. the first stage winners, that we are seeing now are the hardware providers, NVDA etc, the second stage winners will be the hyperscalers, GOOGL, MSFT AMZN and maybe META and a couple of others, not too sure the degree of success here, could vary a lot as ROI varies, then the most interesting phase is customised use cases that probably see the new gorillas emerge with applications on top of the infrastructure. Certainly the hyperscalers will be involved here but there will likely be new entrants. so exciting times ahead, imo. i dont think this is a fad that will pass, but nothing goes up in a straight line so there will be opportunities. i say not a fad because there are too many sensible sources saying benefits are real and large, even at this early stage eg TSM.
as for losers, those areas that can be replicated by AI, call centres first, also some are pointing to the data storage and data manipulators such as SNOW and their competitors. a bit past my pay grade here but seems AI may pass them by. anyone supplying easy to replicate services, mainly labour i would think. the benefits simply reallocate.
at this stage my best bet is the strong get stronger as they (ie as long as they) efficiently embed AI into there processes.
GOOG? That's a contrarian bet.
I would think META but not GOOG. I know Zuck is a seller but Meta is pretty on point with GenAI I think. And also Msft
Maybe elaborate a bit more on GOOG as honestly I haven't looked at them recently since their AI Dev issues.
It's also interesting that NVIDIA has grown so big it can influence market movements like last time when NVIDIA outlook was below analysts expectations and everything exposed to AI fell as a result.
A little left field, but with visions of supplying services to end customers in this space is Data#3 $DTL. They may be a bit difficult to get for an online meeting now that they are in the ASX200 with a $1.3Bn market cap, but who knows, as they are still a Brissie start-up at heart.
Brad Colledge is the MD @Strawman