While we often focus on the short-term metrics of AI and cloud adoption, it's important to pull back occasionally and look at the big picture of the last 10-20 years, as we transitioned toward strong adoption of technology across all parts of our lives. This process helps us qualify what the next tech sales opportunity can look like and whether AI is really the game-changer many believe it to be.

We will focus on the recent 340-page report by Bond Capital—an offshoot of the famous VC Kleiner Perkins that focuses on what they see as "blue-chip" tech companies that already have strong traction in their target markets. This is a selection of the type of companies they invest in, which overlaps with much of our focus here on cloud infrastructure software:

"How to sell AI" v2.0 is now live and discounted for a short time!

Over the last 10 weeks, every week I've added new content covering AI testimonials from Nasqad, Sanofi; deep dives into new technologies like MCP and OpenAI Codex; as well as "state of play" content around ASI and hardware production. As promised, the weekly content will keep going.

This week I cover the risks involved in getting a role in an AI/data startup. The pace of change in the industry is rapid and "occasional updates" won't cut it.

Code WIN2 gets you 20% off to get you started today - it expires today.

The key takeaway

For tech sales: AI adoption is accelerating aggressively in the first half of the year, with reasoning models leading to a significant increase in inference. This shouldn't come as a surprise as we zoom out and look at this moment as part of a buildup of computing usage across the last few decades. The big question is not whether AI matters, but whether your company is able to offer an improved product for your core customer base.

For investors: In the last 4 years, we've seen the spectacular fall from grace of fundamental research, with the large-scale entry of retail investors into the market. Many of the biggest stock outperformers can be linked to retail interest, rather than the company offering exceptional value that was previously not priced in by the market. I think that this will start shifting in the next few years, particularly as we progress toward AGI, because we are going to start seeing more new products and inventions that "break the game" in their specific niche. Self-driving does that for the automotive industry; predictive protein synthesis achieves a similar shift for medical research. The obvious winners here will emerge in several years, but this is the time when those early bets are likely to put you in the right position ahead of the trend becoming obvious.

Slowly at first, then quickly

Source: Trends in Artificial Intelligence by Bond Capital

Trying to measure GDP growth is a difficult task, to say the least, but it is interesting to put into perspective how much growth we've seen in the last 30 years as connected computing (internet, followed by mobile and cloud) became the norm. Depending on the estimates, we are looking at close to doubling the world's GDP in 30% of the time that we've had more advanced technology slowly permeate the economy.

As any new paradigm shift gets introduced, it takes a significant period for costs to go down quickly enough to allow for widespread adoption. AI inference is experiencing dramatically higher cost reduction, and much faster than anything we've seen before.

This is even more important when we account for the overwhelming growth that ChatGPT alone has had. This chart somewhat undersells the opportunity, since the expectation for OpenAI is to end up at between $10B and $15B in revenue by the end of 2025.

We are seeing a similar shift with physical AI, where working products are starting to capture a significant market share.

Structurally, AI is also starting to quickly take over market share from arguably the most important function of modern computing—searching and finding relevant information.

This is an outline of the capabilities that ChatGPT can deliver today..

This could happen potentially over the next 5 years.

This could potentially happen over the next 10 years (many believe it will happen faster due to AGI).

In fact, LLMs are progressing so quickly that they are now starting to become difficult to distinguish from humans.

ChatGPT is the fastest-growing application that has ever debuted on the market. Its pace of user acquisition is unprecedented, particularly when we account for the fact that its usefulness is fundamentally tied to "productivity"—the previous top 5 apps that had such rapid growth were all entertainment-focused.

Exposure to new technologies is a critical path to monetizing them over time. LLMs are slowly crossing through a traditionally difficult gap—the age divide between users. Arguably, multi-modal applications are very well suited for older generations of users because they can just speak to the app and show it things through the camera.

All of this usage will need to be serviced, which leads to significant expansion of the devices capable of processing AI requests (inference). While personal computing and mobile were already extremely successful and ubiquitous, AI is predominantly cloud-based, which means that there are tremendous amounts of servers that will be installed over time to service the constant demand.

Building out that hardware capacity is not an easy task, as CAPEX investments from the hyperscalers have skyrocketed in recent years, causing significant concern among investors. This is a game suitable only for the best-funded players.

Money alone, however, is not sufficient to play. Energy and space constraints have the biggest impact on the ability to scale, both in terms of production and building out those datacenters. Companies are starting to think in terms of "megawatts required to solve a specific problem." Getting access to the grid at the right level means that your datacenter needs to have political support—we are not able to just add that capacity quickly to the grid as a new workload.

The capacity needs will keep increasing as more AI-optimized software is delivered to the market. Since the launch of CUDA, there has been a significant expansion of developers working with the hardware, which is then leading to downstream demand as their applications start generating workloads.

Just in the last 4 years, we've seen an explosion of developers, startups, and applications that are leveraging accelerated computing. This has downstream effects across the whole value chain.

Currently, enterprises are running AI implementations in two general groups—business-generating and cost-cutting. While the cost-cutting value proposition is easier to model before purchase, the really interesting opportunities will stem from revenue generation, and there is increased recognition of that across decision makers.

Executives are pivoting to AI when talking to investors. While this has been perceived mostly cynically, if we start looking at the progress made in a number of S&P 500 companies and the type of transformational changes they are experiencing, this is closer to a lagging indicator of adoption rather than hype.

I talk often about the hyperscalers being the "foundation" that cloud infrastructure software sits on. Particularly with marketplaces, there is a lot of revenue across the ecosystem that passes through their systems. Growth continues to be exceptional, even when we account for smaller organizations that don't really deserve the name "hyperscaler."

Arguably the most interesting part of the report from a tech sales perspective is "what's next" in enterprise adoption. There are a number of potential verticals that are changing rapidly as LLMs and ML are being implemented at scale.

If we look at the field of medicine and scientific research, protein sequencing and structure prediction is experiencing significant milestones in the last few years. The expected outcome of those breakthroughs would be significantly accelerated drug development, personalized medicine and major productivity improvements that free up researchers on more valuable tasks. For additional information, check AlphaFold and the work of DeepMind’s CEO (the primary AI lab for Google), who won a Nobel prize in chemistry for his work in 2024.

Traditional SaaS companies themselves are going through a rapid transformation, with some very interesting pivots in behemoths like Salesforce and SAP. Salesforce fundamentally has refocused toward agentic workflows as part of an extended "digital workforce."

Physical AI will have tremendous economic impact on the world, and it will start with self-driving. Both Tesla's and Waymo's progress in the last 12 months has been staggering, to say the least.

It's important to put into context the fact that AI right now is a two-horse race, both in terms of innovation and usage. Few other countries are able to compete in terms of compute and developers, which results in significant consolidation of the AI opportunity for tech sales within US companies.

We'll close with the report's conclusion:

Imagine, for a moment, how different your next week would look if there were no internet. Every facet of modern life – how we work, how we communicate, how we govern, and more – would likely be turned on its head. The internet has been woven into so many facets of life, big and small, that – for many – it is difficult to imagine a world without it.
In the next decade or two, imagining a world without AI will likely feel the same. Artificial intelligence is reshaping the modern landscape at breakneck speed. What began as research has scaled into emerging core infrastructure across industries – powering everything from customer support to software development, scientific discovery, education, and manufacturing. This document has aimed to map the pace and breadth of AI’s expansion, with particular focus on usage trends, cost dynamics, infrastructure buildout, and early monetization models.
The through-line is clear: AI is accelerating, touching more domains, and becoming more embedded in how work gets done. Catalyzing this growth is the global availability of easy-to-use multimodal AI tools (like ChatGPT) on pervasive mobile devices, augmented by a steep decline in inference costs and an explosion in model availability. Both closed and open-source tools are now widely accessible and increasingly capable, enabling solo developers, startups, and enterprises alike to experiment and deploy with minimal friction.
Meanwhile, large tech incumbents are weaving AI deeper into their products – rolling out copilots, assistants, and even agents that reframe how users engage with technology. Whether through embedded intelligence in SaaS or agentic workflows in consumer apps, the interface layer is being rewritten in real time.
On the compute side, investment continues to scale dramatically. Capital expenditures across major cloud providers, chipmakers, and hyperscalers have hit new highs, driven by the race to enable real-time, high-volume inference at scale. The investment is not just in chips, but also in new data centers, networking infrastructure, and energy systems to support growing demand. Whether this level of capital expenditure persists remains to be seen, but as AI moves closer to the edge – in vehicles, farms, labs, and homes – the distinction between digital and physical infrastructure continues to blur.
The global race to build and deploy frontier AI systems is increasingly defined by the strategic rivalry between the United States and China. While USA companies have led the charge in model innovation, custom silicon, and cloud-scale deployment to-date, China is advancing quickly in open-source development, national infrastructure, and state-backed coordination. Both nations view AI not only as an economic tailwind but also as a lever of geopolitical influence.
These competing AI ecosystems are amplifying the urgency for sovereignty, security, and speed… In this environment, innovation is not just a business advantage; it is national posture.
As Microsoft Vice Chair and President Brad Smith recently noted,
Given the nature of technology markets and their potential network effects, this race between the U.S. and China for international influence likely will be won by the fastest first mover. Hence, the United States needs a smart international strategy to rapidly support American AI around the world…
The Chinese wisely recognize that if a country standardizes on China’s AI platform, it likely will continue to rely on that platform in the future. The best response for the United States is not to complain about the competition but to ensure we win the race ahead. This will require that we move quickly and effectively to promote American AI as a superior alternative. And it will need the involvement and support of American allies and friends.
Lastly, AI is changing how we interact with the world around us. With affordable satellite connectivity expanding access to remote and underserved regions, the next wave of internet users will likely come online through AI-native experiences –skipping traditional app ecosystems and jumping straight into conversational, multimodal agents.
Similarly, AI uptake is accelerating in the workplace and has the potential to shape how people spend the one third of their lives at work. As usage patterns evolve and unit costs decline, we may be witnessing the early stages of an internet where intelligence is the default interface – accessible, contextual, and increasingly personal.
This is all amplified by the growing flow and transparency of information and capital – and the increasing examples of weaponization. It comes at a time when global powers are more openly asserting autocracy-versus-democracy agendas. As technology and geopolitics increasingly intertwine, uncertainty is rising.
One thing is certain – it’s game time for AI, and it’s only getting more intense…and the genie is not going back in the bottle.

The time to start selling AI is NOW.

The Tech Sales Newsletter #96: How AI fits the big picture of computing

The key takeaway

Slowly at first, then quickly

The Tech Sales Newsletter #97: The Deel conspiracy expands

The Tech Sales Newsletter #95: The golden age of silicon