Ai does not go for your work – he is heading for your whole company. He is increasingly undermining the business models of even large corporations
“I haven't seen AI to be so impressive, but it will probably be soon,” noted James Villarrubia, Director of Digital Innovation at NASA CAS in an interview with FastCOMPany. Sean McGregor from Responsible AI Collaborative added that many positions require responsibility for people or dangerous procedures.
“The current AI technology is not yet reliable enough to manage hazardous chemical apparatus or human research without supervision – and there are specialists who are supposed to replace today” – he explains.
However, we do not think ambitiously enough. Hardly anyone points out that artificial intelligence does not have to overcome a single role – it can overcome and re -build the entire system.
Suzanne Rabicofff, founder of Think Tank The Pie Grower, noticed in her analyzes that More than about the exchange of employees, it is about replacing entire, outdated work systems. “Only those more efficient companies will win with AI,” he argues.
Read also at Business Insider
It is enough for AI to be cheaper and faster than complex procedures of a great organization to mix up on the market. In other words, artificial intelligence does not have to be better than every employee – all you have to do is outdo the whole system in which individual specialists work.
Check also: Hundreds of employees from India pretended to be AI. Shocking fall of GWiad
The wave of investment and adoption accelerates
Gartner noted that at the end of 2023 half of the surveyed companies created their own AI tools, but at the end of 2024 this percentage dropped to 20 percent. – The rest put on external suppliers, which increased competitive pressure.
At the same time, AI start-ups attracted a record 46.4 percent. from $ 209 billion global financing of Venture Capital. According to the latest AI Index 2025, already 78 percent companies admitted that they use AI in at least one business function – in 2023 it was 55 percent.
IDC forecasts that enterprise spending on AI solutions will reach $ 307 billion. in 2025, and by 2028 they will increase to $ 632 billion, of which the generative segment itself will exceed $ 202 billion. McKinsey estimates that generative AI can add 2.6 to 4.4 trillion every year. values to the global economy. World Economic Forum indicates that 86 percent global employers expected that AI and broadly understood information processing will “transform” their activities before 2030.
Who is most vulnerable in all this? Just reach for history. In the 1970s, Japanese car manufacturers were the first to introduce robotics, which allowed them to bounce off the Americans – the US's share in the global market fell from 48 percent. in 1960 to 23 percent in 1980
A similar fate today threatens companies with technological debts, including in banking, insurance, telecoms or traditional retail trade. They are characterized by many years of IT systems, complex chain of decision acceptance and expensive bureaucracy. If a smaller competitor (e.g. start-up), strongly using AI, offers the same service cheaper, faster and even in the Self-Service model, Customer loyalty to outdated giants quickly stops.
And large companies often even ask for market interference. In many cases, they work in silos, where one department must strive for the resources of the other department. People who show an innovative approach and want to introduce modern solutions in the corporation are blocked by employees who prefer order and implementation of tasks in accordance with the plan set out a year or even two years earlier.
In corporate environments, an idea for something that can protect the business model or significantly improve it, goes to the so -called Backloga, because first things are realized, which – in the era of extremely rapidly developing technology – are often outdated at the start. Corporations lack agility, and this is one of the most important things to adapt to market changes in time.
Read also: Artificial intelligence incomparably better than a doctor. Surprising research results in the USA
What should large organizations do
Jonathan Rosenberg, former Vice President of Google, He often emphasized that “it is not enough to be great in what you do – you have to catch a big wave”. For the corporation, this wave is artificial intelligence today. Just how to catch her?
There are at least a few changes that large enterprises should introduce, which do not want to be behind agile start-ups and SME companies focusing on modern technologies:
1. Strategic slimming of processes. Each step requiring manual approval slows down the implementation of algorithms. Market leaders create a unified layer of data and APIs that allow you to experiment in weeks, not in years. They also often have so -called Internal start-ups-the environment, where small teams in isolation from the corporation, in a kind of “sandbox”, test the latest business models.
2. Upskilling and the culture of experimenting. Ankur Patel, CEO of Multimodal, encourages experimenting. “AI will not become a team leader, or at least not in the next generation – and this is where there is a great opportunity to develop people.” Leadership competences, integration between departments and change management become an advantage that the algorithm will not duplicate.
3. Zeroing technological debt. Companies that, for example, moved the most important systems to the cloud, reach an average of 30-50 percent. Shorter time of the AI experiments cycle (McKinsey data from 2024). If we see that some tools or systems are clearly slowing us down, we cannot underestimate it.
4. A partner program with start-ups. Venture Client cooperation allows you to test ready -made models or AI agents without the need for many years of implementation projects.
See also: AI is a test for the management. The average cost reduction will be the average
This wave is just accelerating
The cost of launching models with GPT-3.5 efficiency dropped in 18 months from 20 to just $ 0.07. for a million tokens. The scale of acceleration is illustrated by the fact that 60 percent According to IDC, companies from the IDC list will transfer critical loads to infrastructure optimized under AI until 2027.
However, if in three years generative models are another several hundred times cheaper, the difference between the AI start-up and corporation burdened with procedures may become an unacceptable abyss.
The conclusions arise on their own. The labor market will survive, but only where the companies themselves survive. Artificial intelligence does not hunt for individuals – it prescribes the sense of the existence of entire organizations.
They will win those businesses that slimming processes today, invest in data architecture and give people space to consciously use AI and reward innovation and experimenting. They will lose those that – yes – employ great specialists, but attach them to the inelastic system and outdated processes.
The question is not whether AI will take a job, but whether it will allow our company to survive the next decade.
Author: Grzegorz Kubera, Business Insider Polska journalist
