2024-05-10 13:41:48
We now know that we must prepare business stakeholders for Artificial Intelligence, but do we still need to prepare our organizations for it?
The obstacles currently encountered by organizations in deploying AI are numerous. If the fears are human, the response can be organizational and more particularly supported by “AI Governance” formalizing the procedure to follow and intervening in targeted cases as arbiter of decisions.
Preparing organizations for artificial intelligence
AI will quickly require choices that will impact the organization of the company, in particular through AI risk management, compliance with new regulations on AI or even the need to adopt ethical rules for the use of AI. These areas are already well identified and are not major obstacles to its deployment but are required conditions for its proper functioning.
In order to anticipate the organizational obstacles of AI, we need to better understand its impacts and to do this, return to our 3 families of AI: targeted AI, integrated AI and interactive AI (3.1) . The table below summarizes the main organizational impacts to anticipate:
| Type of decision | Impact of the decision on the rest of the organization | Organizational impacts | |
| Targeted AI | Existing | Non | None |
| Integrated AI | New | No/limited | Process optimization |
| IA interactives | New | Oui | Complete overhaul of organizations & processes Vertical integration Creation of new entities |
By definition, targeted AIs have no impact on the organizationunless we encourage the move to scale and as we have seen (2.1), the move to scale is itself confronted with its own limits.
Integrated AI, for its part, will have impacts on the organization – essentially process optimizations – with human interventions (the concept of “ human-in-the-loop » is well assimilated by AI players) when AI is deployed on high-stakes subjects. Furthermore, AI integrated into products will help to redefine R&D or product marketing in many industries, starting with mobility as is already the case today.
He is more difficult to anticipate the organizational impacts of interactive AI. However, in a world where AI is optimized globally and no longer locally, the possibilities are endless. I see three at this stage that many companies will not escape:
- A complete overhaul of the organization and processes induced by a concentration of decisions from AIs, to the extent that the different AIs respond to a common objective, they must coordinate
- Vertical integration, particularly in the area of the supply chain where joint decisions will improve the results of separate entities on the value chain
- Finally, the emergence of new entities, in joint venture mode, associated to build services or products whose skills are located in a multiplicity of organizations but whose operational decisions associated with AI will be taken by a global AI .
These possibilities, theoretical at this stage, obviously open up questions regarding the evolution of major economic sectors, the distribution of added value between the different market players or even the right to competition, but obviously other developments, similar or opposed, will materialize on a market which risks evolving very quickly…
In view of the challenges ahead, it is clear that future transformations will have to be carried out by business management. They will be carried out within companies but also beyond its walls…
And beyond the company?
The impacts of AI for European companies cannot be decided only at their level. They evolve in an environment that exceeds their limits. Today, the competitiveness of companies on AI is decided both within them, at the European Commission in Brussels, but also on school benches, from primary school to university.
The case of the GDPR which places all the costs and risks on European companies, and restricts the ability of companies to rely on part of their data, this observation observed by most practitioners has just been confirmed by an excellent study entitled “ Data Privacy laws and firm production : evidence from the GDPR » recently carried out by four economists. The observation is unequivocal: since the entry into force of the GDPR, European companies collect and store 26% less data than their counterparts across the Atlantic in comparison with the previous period, they also process 15% less of this data. The impact on the competitiveness of our companies in a Data and AI economy is already perceptible.
In this context, the 15th recommendation of the recent report of the Commission on Artificial Intelligence which aims to “transform our approach to personal data to protect while facilitating innovation to serve our needs” pleads for the relaxation of the application of the GDPR by the CNIL. We cannot underestimate the importance of relying on data – in volume and quality – to build efficient AI. The report also recommends opening up health data (10th recommendation). This recommendation must extend to other sectors of the economy and, starting with electricity and in particular for electricity networks, or even finance, in particular on sustainable finance and the fight once morest money laundering and financing. of terrorism.
This article is part of a series of articles proposed by Patrick Darmon. Why and how AI must become a core technology for our businesses ?
First part: Can artificial intelligence be used in business?
Second part: Artificial Intelligence revolution… or just a dream?
Third part: AI, how much does it cost and is it worth it?
Fourth part: What strategy should you adopt to integrate generative AI into your business?
Fifth part: Don’t scale AI! Scale your business with AI
Sixth part: If you are wondering where to put AI in your business, we have the beginnings of an answer
Patrick Darmon is a partner at Fizz venturecabinet de council en data & IA.
1715558151
#Artificial #intelligence #economic #national #imperative