AI Platforms vs. Hyperscalers

Written By Joerg Rheinboldt

Joerg, 8.1.2026

A HEARTFELT Selection Lens on Agency, Timing, and Founder Quality

Executive Summary

We observe a structural shift in how very early-stage companies are built. In the first months of a company’s life, AI platforms increasingly matter more than hyperscalers. Founders optimize first for intelligence leverage rather than infrastructure leverage. Hyperscalers remain present, but largely as an invisible substrate.

This shift can be turned into a selection heuristic:
A new company should either work AI-platform-first by default—or have a very good, well-articulated reason for doing something else.

How a team positions itself relative to this default is a strong signal for agency, timing awareness, and founder quality.

The Hypothesis

Hypothesis:
Early-stage startups today rely more on AI platforms than on hyperscalers in their first months of growth. They spend more—financially and cognitively—on AI capabilities than on classical cloud infrastructure.

This is not a claim that hyperscalers are irrelevant. Rather:

  • Hyperscalers are increasingly abstracted away.

  • AI platforms are the felt dependency in the earliest phase.

  • Founders optimize for intelligence density before they optimize for infrastructure efficiency.

This hypothesis is most visible in pre-seed and early seed teams, especially:

  • very small founding teams

  • AI-native products

  • software-first B2B companies

Findings: What We Actually See

The Shift in Early Leverage

In the early months, startups typically do not need:

  • sophisticated cloud architectures

  • global redundancy

  • fine-grained cost optimization

They do need:

  • reasoning power

  • language, vision, and coding capabilities

  • fast iteration on product meaning

  • leverage that substitutes for early hires

AI platforms deliver this directly. A small monthly AI spend can replace meaningful human effort. Infrastructure spend remains low and relatively flat early on.

The result: AI becomes the first real burn line.

Cognitive Gravity Matters

Founders increasingly think in terms of:

  • prompts

  • agents

  • evals

  • workflows

Not in terms of:

  • instances

  • regions

  • load balancers

Even when a product technically runs on a hyperscaler, founders often do not experience that relationship. They experience the AI platform. This cognitive gravity matters because dependency precedes lock-in.

Hyperscalers Are Still There (But Backgrounded)

Most startups are implicitly on hyperscalers from day one—via hosting, dev platforms, or databases. But hyperscalers are no longer the primary object of attention early on.

A useful mental stack:

AI Platform → Dev Platform → Hyperscaler → Physics

Founders optimize the top of this stack first.

A Selection Heuristic Emerges

From this observation, a practical HEARTFELT heuristic emerges:

A new company should either be AI-platform-first by default, or be able to clearly explain why it is not.

The explanation itself becomes signal.

This leads to three archetypes we see in selection.

The Three Archetypes

“From the Past” (Common, Negative Signal)

These teams:

  • treat AI as a feature rather than a working substrate

  • default to classical SaaS and infrastructure thinking

  • focus on architecture before leverage

What this usually indicates:

  • mental models shaped by a previous cycle

  • AI adoption without internalization

  • slower iteration and higher burn for equivalent output

This is rarely a contrarian position. It is usually a timing mismatch.

“Default Present” (Expected Baseline)

These teams:

  • use AI platforms aggressively early on

  • substitute cognition before infrastructure

  • focus on workflows, leverage, and speed

This is no longer exceptional. It is the new baseline of competence.

Not a reason to invest—but a reason not to worry.

“Different Universe” (Rare, Strong Positive Signal)

These teams do not go AI-platform-first, but:

  • understand the default deeply

  • reject it consciously

  • can articulate precise trade-offs

Legitimate reasons include:

  • hard regulatory constraints

  • data gravity dominating intelligence leverage

  • edge or on-device requirements

  • infra-level or physics-constrained problems

This category is rare—and often correlates with founders who are early for the next cycle rather than late for the last one.

Connection to HEARTFELT Core Ideas

Agency

Agency, in HEARTFELT terms, is the ability to act effectively within the constraints of the moment.

AI-platform-first behavior often signals:

  • high personal leverage

  • willingness to rethink how work is done

  • active rather than inherited operating models

Teams “from the past” often show procedural agency but lack strategic agency.

Timing

This heuristic is fundamentally about timing.

AI platforms represent the current leverage frontier. Founders who are native to this frontier show:

  • sensitivity to where value is created now

  • ability to update mental models

  • alignment with current opportunity structures

Teams misaligned here are often not wrong in principle—just late.

Founder Quality

The key signal is not which tools founders use, but:

  • whether they understand the default

  • whether they can reason about trade-offs

  • whether they can explain their position clearly

Strong founders are either:

  • deeply fluent in the present, or

  • consciously and convincingly in rebellion against it

Anything else is inertia.

Practical Application in Selection

We should not encode this as a rigid rule. Instead, we should use it as a diagnostic lens.

A useful question in selection:

“In your first 6–12 months, where do you expect most of your leverage to come from—and why?”

We listen less for the answer, and more for:

  • clarity of reasoning

  • awareness of alternatives

  • explicit trade-offs

The quality of thinking is the signal.

Important Caveat

This heuristic must be time-versioned.

AI-platform-first is the right default now. It may not be in five years. The enduring principle is:

Founders should be native to the leverage of their moment—or consciously opposed to it.

That principle, not the tools themselves, is what we should anchor in HEARTFELT’s selection philosophy.

Closing

This lens helps us distinguish:

  • competence vs. inertia

  • contrarian insight vs. being late

  • founders with agency from founders with habits

It is simple, explainable, and grounded in how companies are actually being built today. That makes it valuable—not as doctrine, but as orientation.


Joerg Rheinboldt https://joergrheinboldt.com
Previous

Letters, Love, and the Investor’s Eye