Data Points

Subject: AI ROI: The Emperor Has No Clothes (But Anthropic Does)

Friends,

Let's cut the BS. This week's deluge of "AI ROI" articles is mostly smoke and mirrors. Everyone's patting themselves on the back about pilot projects and incremental efficiency gains. But the real story isn't in the press releases; it's in the funding.

Specifically, look at Anthropic.

The Headline: Anthropic is securing over $10 billion in new funding, potentially exceeding $20 billion when combined with previous commitments from Nvidia and Microsoft. Their revenue run rate more than doubled since mid-2025, reaching over $9 billion at the end of last year.

The Subtext (and what everyone's missing): That's ROI. Not some nebulous "improved customer experience" metric. Real, hard cash flowing into and out of a company building AI.

Here's why this matters:

The vast majority of "AI ROI" projects are focused on optimizing existing processes. Think: chatbots for customer service, AI-powered fraud detection, or automated report generation. These are all fine, but they're marginal improvements. They're not creating entirely new markets or fundamentally changing how businesses operate.

Anthropic, on the other hand, is building the infrastructure upon which these applications are built. They're creating the shovels in the gold rush. And investors are betting big that those shovels will be worth a fortune.

My Hot Take: Most companies are approaching AI ROI backwards. They're starting with the application and trying to justify the investment. They should be starting with the infrastructure and asking, "What becomes possible when we have access to truly intelligent machines?"

The Problem: This requires a completely different mindset. It requires:

• Long-term vision: You're not looking for immediate payback. You're building for the future.
• Risk tolerance: Foundational AI research is inherently risky. There's no guarantee of success.
• Talent: You need people who understand both the technology and the business implications. Think someone like Daniela Amodei, CEO of Anthropic.

Advice for you:

1. Stop chasing incremental gains. Challenge your team to think bigger. What moonshots are possible with AI?
2. Invest in AI literacy. Train your people to understand the capabilities and limitations of AI.
3. Don't be afraid to experiment. Start small, but think big. Allocate a portion of your budget to high-risk, high-reward AI projects.
4. Follow the money. Pay attention to where the smart money is flowing. Anthropic is a good place to start.

The future belongs to those who build the infrastructure, not those who simply optimize the existing system.

Stay sharp,

Yves Mulkers
Ins7ghts

Subject: RE: AI ROI: The Emperor Has No Clothes (But Anthropic Does)

Friends,

Let's cut the BS. This week's deluge of "AI ROI" articles is mostly smoke and mirrors. Everyone's patting themselves on the back for marginal gains while ignoring the elephant in the room: most AI projects are still cost centers, not profit centers.

And to answer the question of which data supports your client's business case... well, that depends on the narrative you want to sell. Because the actual data is messy.

Here's the uncomfortable truth:

The "ROI" numbers being thrown around are often based on cherry-picked use cases and questionable attribution. Sure, a chatbot might* have reduced customer service costs by 10%, but what about the cost of training, maintenance, and the inevitable PR disaster when it goes off the rails?
Everyone's focusing on the easy* problems. Automating simple tasks is great, but it's not going to fundamentally transform your business. The real ROI comes from tackling complex, strategic challenges – and that's where most AI projects fall flat.

• The talent gap is real, and it's killing ROI. You can't just throw AI at a problem and expect it to solve itself. You need skilled data scientists, engineers, and domain experts to build, deploy, and maintain these systems. And those people are expensive.

However, there are glimmers of hope. And they all point to one thing: focus on infrastructure, not applications.

Look at Anthropic. While everyone else is arguing about the ROI of specific AI applications, they're quietly building a massive AI computing infrastructure. They just secured over $10 billion in funding, bringing their total to over $20 billion. That's not because they're building the best chatbot; it's because they're building the picks and shovels for the AI gold rush.

My hot take: The real AI ROI isn't in the applications; it's in the enabling infrastructure. If you can build a robust, scalable, and secure AI platform, you can unlock a whole range of applications and drive real business value.

And speaking of infrastructure, keep an eye on developments like Dynamic Differential Linear Attention (DyDiLA). While the name sounds like something out of a sci-fi movie, it promises to improve the efficiency and quality of linear diffusion transformers. That means faster, cheaper AI model training – a key enabler of ROI.

Advice for your client:

1. Stop chasing the shiny objects. Don't get distracted by the latest AI hype. Focus on building a solid data foundation and a scalable AI infrastructure.
2. Invest in talent. Hire (or train) people who understand both the technology and the business.
3. Think long-term. AI is not a quick fix. It's a strategic investment that requires patience, persistence, and a willingness to experiment.

And finally, be prepared for regulatory headwinds. 2026 is shaping up to be a battleground between federal deregulation and state-level AI rulemaking. That could significantly impact the cost and complexity of AI projects.

Until next week,

Yves

Subject: RE: AI ROI: The Emperor Has No Clothes (But Anthropic Does)

Friends,

Let's cut the BS. This week's deluge of "AI ROI" articles is mostly smoke and mirrors. Everyone's patting themselves on the back for "AI ROI" while conveniently forgetting to mention which AI and at what cost.

You asked about the source. The real source is desperation. Desperation to justify inflated valuations, desperate to prove to boards that these massive "AI investments" are actually paying off.

But let's get real. Most of what's being touted as ROI is just cost displacement disguised as innovation. Slashing headcount and calling it "AI-driven efficiency" doesn't magically create value. It just shifts the problem (and often degrades the customer experience).

Here's the breakdown of what I'm seeing, and what actually matters:

The Anthropic Exception:

Now, about Anthropic. While everyone's distracted by OpenAI's latest shiny object (ChatGPT Health, apparently - see FayObserver), Anthropic is quietly building a real business. Their focus on safety and explainability is resonating with enterprises in regulated industries. I'm seeing actual, measurable ROI in areas like:

• Compliance: Claude is being used to automate compliance checks and generate audit reports, saving significant time and resources.
• Risk Management: Anthropic's models are better at identifying and mitigating risks in complex financial transactions.

The Apple/Google Paradox:

The news of Apple partnering with Google (iClarified) is interesting, but it highlights a key problem: even the biggest tech companies are struggling to build truly differentiated AI. Apple needs Google's AI muscle to stay competitive, but at what cost to their brand and user privacy?

My Hot Take:

The real ROI in AI isn't about replacing humans; it's about augmenting them. Focus on use cases where AI can handle the mundane, repetitive tasks, freeing up human experts to focus on higher-value work. Stop chasing the hype and start focusing on practical applications that deliver tangible results.

Advice for Your Client:

Don't buy the hype. Demand specific, measurable results. Focus on use cases where AI can augment human capabilities, not replace them entirely. And seriously consider Anthropic. Their focus on safety and explainability is a major differentiator in today's market.

Until next week,

Yves