It is easy to buy AI but it is hard to make it profitable. Recently, I have noticed that a lot of companies subscribe to AI services and activate virtual assistants because of the force of joining the trend. Everything seems great before the end of the quarter and coming invoices. With all that investment, why nothing appears in the income statement? [1]

In my opinion, the problem is not the technology itself. It is the financial naivety. AI is not facing an adoption problem but a constant problem of integration.

1. Companies acquired a tool but did not redesign the process

In my opinion, companies are treating the AI tools the same way they treat other software layers. Something that you add on top of existing systems and expect the productivity growth to happen automatically. It will not work.

For example:

  • If customer support answers emails faster but still takes several days to resolve the client’s issue due to the internal bottlenecks, then we only sped up the bureaucracy.
  • If analysts can generate ten financial reports within an afternoon but we did not redesign the decision-making process, then we only generated more PDF reports.

For me, this is the main error: applying AI without changing the overall process.

2. Cost becomes obvious but the value stays vague

Now we observe that companies spend millions of dollars on tokens within one month. [2][3] As the result, companies start to set up limits, push cheaper models and try to control the token consumption [2]. It is important for me since it breaks the myth: the myth that you can adopt first and somehow numbers will sort themselves out later.

No. Numbers are already here and they come in a large number.

The problem is that for many companies the cost becomes more and more obvious, but the value from the client’s perspective remains obscure. There is more activity, more tests, more tokens, more internal presentations. But there may be no more margins, revenue, or operational benefits [1].

3. Companies continue to invest despite everything because nobody wants to be left behind

Companies continue to invest with two hands and it is surprising for me.

They are not doing this because they start seeing the return. In my opinion, they are doing this because they fear being left behind. If the AI will redefine the way of work, selling, or company operations, nobody wants to be the one that reacted two years too late. (Hello, Apple.)

That is why it explains the current paradox quite well: investment grows rapidly but the monetization becomes blurred [4]. It is not a sign of success. This is called FOMO [5].

Conclusion

My point of view right now is that companies do not lose money with AI because of the lack of its efficiency. They lose money because they try to graft it into existing processes hoping to achieve new results.

As long as the AI is introduced into companies not as the redesign of processes but as a separate tool, it will create more motion than value. This is why even as the adoption is increasing the profits will remain to come later.

References