That is why, so as to spark what we hope will be an important debate, we imagined 4 scenarios of economic organization for the age of AI. Sometimes, they are complementary; other times, antagonistic. Fast forward to 2030.
What is the most beautiful love story of the 21st century? It is the encounter between AI and the blockchain. The creation of a company or association used to be reserved for a small number who dared to take the plunge; today, few are those who have not founded several organizations during their lifetime. By-laws creation, registration, accounting, relations with the administration … everything is done automatically.
But the encounter between AI and blockchain has changed above all the very definition of what is the sacrosanct « venture »: as it is very easy to measure one’s participation in a project (the impact that the brand name proposal, the answer to some users on a forum or the development of a new feature has had on the popularity of a service), the founder / contributor distinction no longer makes any sense – as for the employee / volunteer dichotomy. There is only a long continuum of involvement, the various levels of which are rewarded by more or less crypto coins related to the projects. On average, an adult is active on 7 different projects.
No one is willing to work full-time on a single project, as the efficiency gains derived from AI have lowered the legal working time to 20 hours per week. Since individual projects are numerous and henceforth more time-consuming than employment narrowly defined, the inclusion in a smart employment contract of an exclusivity clause has been illegal for 6 years now.
The outsourcing trend, which entails companies focusing on their core business (even if the definition of what constitutes a core business is changing quite often) and transferring certain activities (e.g. payroll, IT) to external providers, has multiplied exponentially since the irruption of artificial intelligence.
Because the scramble for large volumes of data and the need to master a very specific problem (e.g. predictive detection of technical problems on industrial engines) naturally leads to a proliferation of highly specialized actors, proposing narrow AI algorithms. It is useless to reinvent the wheel in each firm when a third-party actor can aggregate problem-specific data and thus create a better algorithm that will serve everyone.
Every company therefore uses thousands of APIs that allow it to exploit the of best collective intelligence on every subject. An API of advertising spending decisions, another of translation of UIs in 200 languages, another one of rental and management of delivery robots depending on demand… AI APIs orchestrate the movement of the atoms just as they had conquered bits.
And then AI has this magic touch that it allows to automatically find the best resources for each project. One plugs her favorite search algorithm on one of the 3 major API marketplaces (among the most powerful groups in the world), types in her request, the algorithm chooses the best API for the problem described, before the algorithm-architect of the information system takes over for the technical integration (in the background: the user never has to put the hands in the sludge, or it is by nostalgia for GitHub, Linux and Amazon Web Services).
Nowadays, apart from a few technical employees (1% of headcount), every teammate has the freedom not to ever worry about the tedious technical execution of projects. It is the generation of ideas and the design of new services that really matter. The consequence being a Cambrian explosion of the number of innovations generated every year.
In the era of artificial intelligence, data network effects (more data => better service => more users => more data etc.) are so powerful that the centripetal force of business concentration hits records in the light of the mankind’s economic history. As bits of information on clients are miscible (you can combine a bank statement with a Pinterest board to track preferences), and the marginal value of a bit is increasing (the more data I already own about a client, the better I will be able to exploit each additional data point), the boundaries between industries have irremediably dissolved. The race to scale is further amplified by international competition. In 2030, to acquire or to be acquired, that is the question – a burning one.
After taking over Amgen, Nike and Coca-Cola, Amazon has just set its sights on American Airlines. The old glories of the industrial age are not to be outdone: General Electric (which has recently to manage 3 European countries) has announced the biggest merger ever seen, an alliance with Tencent and Allianz. And (financial) rumor has it that Teslook and Danone could be joining forces soon.
Business gigantism leads to an equally impressive war on the talent side. The world has not trained enough AI specialists despite the take-off of demand, and the self-catalytic effects of machine learning have increased their productivity tenfold. Inevitable result: their market value has soared, as has their aura. (So much so that the neural stream « Deep Gossip » now exceeds the Daily Mail in daily connections.)
The conglomerate model is thus the following:
- Welcoming the most promising scientists.
- Training them at their in-house universities, which have overtaken for a long time the prestigious institutions of the past (Stanford is now considered second-class).
- Making them work on the most general possible problems, the ones that are profitable in any sector.
- And reinvesting the profits generated in recruitment, training and, above all, retention of these super-stars.
For instance, Alphabet just announced the recruitment of a 12-year-old Ugandan prodigy identified in a planetary competition combining mathematics and programming (after validation from an economic valuation algorithm gauging her cognitive potential). She is now joining Mountain View’s training center with a record salary of €20m per year, with a commitment to stay at least 30 years. Any early termination will result in the automatic repayment of the sums previously received.
The concept of « matrix organization » makes the avatars of VBS (virtual business schools) classes laugh a lot. A structure capable of deploying itself in only 2 dimensions (by country and function) is as outdated for a current student as could be a silent movie for a Millennial.
Indeed, AI has crowned VDO (Variable-Dimensions Organization) as the only model of an economically viable business structure. Organizational agility is no longer a quality but a necessity, at a time of unbridled pace of new products and services introduction on the true Single Market – that of global scale.
VDO is a fundamental transformation, offering a joint solution to two main challenges:
- Enabling present-day businesses to operate on more dimensions than the poor 2 to which their 20th century ancestors were confined. It is not uncommon for an employee to be linked to more than 5 groups: customer relationship, product, main skill, technology, work habits, etc. These dimensions are more often oriented to be « means » of the employee than « ends » of the company.
- Allowing companies to transform on an ongoing basis. The organization must no longer be a limiting factor to projects; on the contrary, it must embrace their constraints. In general, a dimension is profoundly reshaped every 8 months – rearrangement of its subdivisions.
AI made VDO possible, through the upheaval of the manager-employee microeconomics. Previously, a manager could only effectively manage up to 7 or 8 employees. The growth in size of companies had logically led to the construction of ever-higher organizational pyramids. Everything has changed with smart assistants, automatic translation, algorithms for optimizing the composition of teams, clone bots (customized software decals, capable of responding to the simplest queries as the original would do…). A manager can now accompany more than 60 people, and conversely each employee can report to her managers in each of the divisions.
An explosion of connections among colleagues that makes VDOs shifting ecosystems capable of reaching tremendous scale while flattening hierarchical strata.
Imagine the future of your organization with our AI LabContact us