Where are you seeing the activity?
To write checks to applications in general to buy?
Yeah, just to be technical for a moment, a lot of that is driven by the fact that
these foundation models are translating text into other text.
Well, so what you're seeing is that

the money is flowing into places where
applications can do well with text, as you can imagine.
So that's really the goal tech. It's health care.
It's fintech. And largely that's because the
architectures are settling in terms of

how developers build applications in
2023. What they're still trying to figure out
how to build applications using generative AI.
That has changed greatly in 2024 as applications move into more generic
architectures. What's been so interesting is, for
example, you're an investor in

Anthropic.
We announced today that Amazon is taking another $4 billion cash infusion into
the business. It really wants to make it a bedrock of
bedrock. It wants to really ensure that it's
offering its own worm cloud technology,

but also to other clients with the help
of Anthropic, but also having Anthropic train its own data on its own
chips. What's interesting, I'm talking to some
application generative AI application makers and they're saying the support
coming from cloud providers for their

businesses too, is extraordinary.
How are you seeing these application makers also dining out from Amazon and
Microsoft in the rivalry there? But it's really, really expensive to
train and provide insurance for these things.
You know, seeing a $4 billion cash

infusion from the likes of Amazon is not
shocking at all. I think it's great for developers who
sit in the Amazon ecosystem. They've already been blessed to use
services within that ecosystem and to move over to Anthropic, I think makes
things really, really easy for them.

And Anthropic is obviously benefiting
from that, just from the research that you've done.
Are we seeing money stop going towards these foundational layers and instead
finding a new home in the application makers?
Or are people still writing checks in both directions?
It's just very different valuations.

Yeah, it's going in both directions,
though. You saw in the report the amount of
spending on foundation models went up 6.5 X from a billion in 2023 to 6.5 billion in
2024. And that's a huge increase.
And then as you mentioned earlier, we

saw the amount of spend in the
application layer go up about eight X to about $4.6 billion in 2024.
So the dollars are flowing in both directions, frankly.
And you mentioned some of the sectors where this makes most obvious sense.
And we've had a lot of these these

startup founders on the show that are
particularly helping serve the legal industry, the healthcare industry.
How much money are you starting to see these ultimate end clients purchasing?
How much are we seeing product market fit?
How much are we actually seeing clients spend money on these products?
I think that we call it PMF in the

venture industry, a product market fit
that, you know, it's definitely there, right?
I you know, I have a legal text tech startup called Eve that's seeing huge,
huge benefit from plaintiffs' side case management.
And the throughput is definitely there.

So, you know, plaintiff-side lawyers are
getting the benefit and they're spending the dollars.
And you know, like you asked earlier, the product-market fit is certainly
there right now and people are buying up.
It's serving up software left and right.

So we're really excited.
But everyone keeps on questioning, return on investment. All these
companies in healthcare, in the legal services, finding that it's productivity
that gets boosted. What else is giving them the return?
Oh, absolutely.

I mean, they're not going to be buying
the software for not getting the value.
So, you know, we hear it anecdotally from folks that they're getting tons and
tons of value out of the software and the spend is just increasing rapidly.
So you can imagine that the value

certainly there pushes us forward because
where is the next vanguard of money to be made within the hype cycle of AI
reality? Tim, Where else are you looking to put
money to work? Yeah.
I mean, it's really focused around the

application there.
Right now I think architecture is, like I mentioned earlier, are settled in
because the architectures are settled, makes it easier to build the
applications. And so you're going to see the velocity
of applications come out much, much more

quickly.
So, you know, I think legal definitely workflow automation,
customer support is certainly a space, marketing, human resources.
So I think I think really any vertical right now is, is ripe for disruption for
AI right now.

Following the AI Money Trail

Welcome to the Bloomberg Technology YouTube channel. Subscribe for original series and mini-docs on game-changing technologies, profiles on innovators, live conversations with experts, and breaking technology news & analysis from Bloomberg TV.

Bloomberg Technology

Investment Trends in AI: Focus on Text-Based Applications

Current investments are concentrated in applications leveraging foundation models for text processing.

Key sectors attracting investment include tech, healthcare, and fintech due to their suitability for text-based AI applications.

The development of applications using generative AI is evolving, with a shift towards more generic architectures.

Significant Investment in Anthropic and Cloud Provider Support

Amazon's $4 billion investment in Anthropic highlights the strategic importance of AI infrastructure.

Cloud providers offer substantial support to application makers using generative AI.

High costs associated with training and insuring AI models are noted.

Investment Distribution: Foundation Models vs. Applications

Investments are flowing into both foundational models and application layers.

Spending on foundation models increased significantly (6.5x) in 2024, reaching $6.5 billion.

Spending on application layers also increased substantially (8x) to approximately $4.6 billion in 2024.

Strong product-market fit is observed in sectors like legal and healthcare.

A legal tech startup, Eve, demonstrates significant benefits and client spending in plaintiff-side case management.

Clients are actively purchasing software, indicating strong value proposition and increasing spend.

Return on Investment and Future Investment Opportunities

High value and rapid increase in spending suggest strong ROI for clients.

Future investment focus is on applications, particularly in areas with established architectures.

Sectors ripe for AI disruption include legal workflow automation, customer support, marketing, and human resources.