It is not about technology, Watson, it is all about use cases!
In the AI gold rush, there are definite haves and have nots. Some “AI” companies are killing it and growing rapidly – at astonishing rates, in valuation, but more importantly in terms of customers, revenues and eventually profits. They are actively hiring (see the chart below) and creating opportunities for people at all stages of their careers. Thinking of the retort that they are not profitable as yet, think of the years Amazon continuously lost money, as it kept investing into its cloud infrastructure. Eventually that investment paid off handsomely.
Employees (June 2024)
Anduril Industries 3500
Anthropic 240
Boston Consulting Group (AI Consulting) 750
CoreWeave Inc. 550
Databricks 8050
Google DeepMind 2000
McKinsey & Company (QuantumBlack) 1000
Palantir Technologies 3936
Snowflake Inc. 7000
Employees (June 2025)
Anduril Industries 4800
Anthropic 1300
Boston Consulting Group (AI Consulting) 1000
CoreWeave Inc. 1450
Databricks 9000
Google DeepMind 6000
McKinsey & Company (QuantumBlack) 1500
Palantir Technologies 4900
Snowflake Inc. 7834
As a potential customer or even an investor – one wonders which company should I pick? Who is going to be the winner and who is going to go out of business as quickly as they raise large sums of money? How do you differentiate between them?
In my opinion, it’s only 2 words – USE CASES.
Let me explain.
All technology provides leverage, mostly from human effort by automating and streamlining processes and functions. Most, if not Al companies and organizations leverage technology to run their operations. So, all of them are operating and are functional.
So why the crazy rush to implement AI?
Mainly because it does more intelligent tasks that till recently only humans could accomplish. However, this is a broad and bold promise. While companies are keen on implementing AI, they are also skeptical.
The main issue most companies are facing are – they cannot ignore AI and its capabilities as it would put them at a competitive disadvantage, however, they do not want their initial project to be unsuccessful as that would de-energize them, and put them at an even further disadvantage.
In my opinion, it boils down to use cases – specific functions or tasks, in a specific domain. Many companies have utilized AI to solve for those tasks, moving more of the work from the human to the “intelligent” software and hence creating scale, automation and even security in a business function.
Some of the use cases can be broken down into a few broad categories as follows:
Gaining or structuring information from documents and data in general – where humans may take a lot of time, current AI can do it quickly and efficiently.
Remembering to get things done, whether it is to follow up on accounts receivables, or to flag a compliance issue and follow up – again AI is more suitable for those tasks.
Doing routine tasks repeatedly – monotonous, but important – like reconciling a portfolio, or bank balances or driving a car or robot, or filling in information that is available.
Processing a lot of information quickly for time sensitive purposes – whether it is credit card fraud detection, or quant trading – AI algorithms make possible what humans and simple software cannot do, specially building intelligence around synthesizing the information and taking steps.
Mainly because AI can do tasks that until recently only humans could accomplish – and in many specific Use Cases, do it better and faster.
There are many, many more such use cases, however, I will submit that AI companies that come with domain expertise and focus on a specific (or a specific set) of Use Cases and using the relevant AI technologies, whether it is agentic ai, large language models, machine learning, natural language processing etc., are the ones gaining the most commercial traction.
As you navigate through the AI journey, my advice to firms looking to implement AI in their business is to focus on a specific use case, implement the(AI)solution, and then move on to the next USE CASE! Build your AI muscle as there is a lot to come.
