2024-03-22 13:54:38
Should we, can we, and finally how to put AI into our business model? Good questions! Patrick Darmon, partner at Fizz venture, will try to answer them in a series of articles of which we are providing you with the first part here.
“The future is already here – it’s just not evenly distributed.” William Gibson, American science fiction writer and one of the leaders of the cyberpunk movement.
February 1914, Henri Ford switched its automobile production line to electric energy a few months after its launch. Electricity has been on industrialists’ minds since 1882, when Edison opened the first power plant on Pearl Street in Manhattan to light offices in the Financial District. Ford is neither the inventor of the production line, inspired by the chains used in slaughterhouses at the beginning of the 20th century, nor the first to use electricity for production – the Ball brothers preceded it by 15 years in their factory manufacturing packaging glass. However, it is indeed Henri Ford who, rightly, is recognized as the inventor of mass production through his innovation which combines a new technology – electricity which operates the assembly line with its 32,000 machine tools allowing to produce the famous Ford T in its Highland Park factory – with managerial innovations such as the vertical integration of suppliers, the increase in the salaries of qualified workers in order to compensate for the arduousness of the work but also to ensure the accessibility of the Ford T to the greatest number, including its workers.
February 2024artificial intelligence is on everyone’s mind as the new General Purpose Technology – the term that economists use to designate the technologies that underpin economic growth through their capacity to multiply and combine innovations and transform all sectors of activity. As in its time for electricity, AI already has its “Edison” on the one hand – the innovators whose Demis Hassabis, Yann Le Cun or even Sam Altman – and on the other hand, industrialists, like Google, Facebook or OpenAI (Edison having been both the genius inventor and the founder of General Electric. Data, the Cloud and algorithms are the news ” utilitys” of this economy. But where are the Henri Fords of AI ? The “100 in Artificial Intelligence” list Time Magazine does not include a captain of industry outside the High-Tech sector! In a period of technological innovation, the High-Tech tropism of the list is understandable, nevertheless a few industrial innovators deserve to be included – even if it is clear that to date, only a handful of them could be there. pretend…
Artificial intelligence is a two-speed highway
Today, artificial intelligence is progressing on a two-lane highway… and at two speeds. On the fast lane, we meet GAFAM, BATX as well as a multitude of new players like Tesla, OpenAI and a few other Tech leaders. These companies are not only at the forefront of AI research – via their numerous research centers and their ability to recruit the best students from the largest universities – but also, at the heart of the new business models made possible by Data, the Cloud and AI.
Other companies, regardless of their size, take the other route. If they sincerely aspire to become “Tech Companies” or other “Data Driven Companies”, we cannot expect them to follow Mark Zuckerberg’s famous “move fast and break things” or even to deliver new Data Products at the pace of a Google or a Tencent. It is, however, clear that they view AI as an important lever, which results in numerous initiatives, for the most part still exploratory and for some of them already operational, particularly in the sectors of health, telecoms, insurance and even industry. Despite this, the general feeling, as expressed by a Chief Data Officer (CDO) in a cutting-edge industry “today AI contributes to results but it is not a game changer “. This makes us wonder what companies should do to move up a gear?
The real challenges of AI in Business
If companies want to enter the artificial intelligence revolution head-on, the approaches followed by the majority of them – from bottom-up collection of use cases to “scaling up AI” – will not be enough. In the age of ChatGPT putting AI in the hands of millions of users, there is a need to take an ambitious approach and turn it towards business transformation. To do this, the future Henri Ford of artificial intelligence will have to respond to two major challenges:
- Focusing the adoption of AI by Business Departments with a view to business transformation
- Remove human, organizational and institutional obstacles, whether internal or external
By posing these issues as major, I am well aware that they could surprise certain CDOs whose agenda is centered on scaling up, so I will clarify their content.
Focusing the adoption of AI by Business Departments with a view to business transformation, this means that AI is becoming a key competence of the company and that for this, we must go well beyond the traditional Bottom-up compilation of business use cases and its roadmap which, too often , leads a life parallel to the company’s strategy and its major operational challenges. This exercise generally results in a list of heterogeneous Data use cases which mixes Business Intelligence, Analytics and Artificial Intelligence and whose impact on the results is not always readable. This approach, which is useful for good appropriation of AI in the take-off phase, is generally not suitable for use on the scale of this technology. Henri Ford would be unknown today if he had focused on use cases for replacing candles with light bulbs in offices rather than looking at the production line. To go beyond this approach, it is necessary to mobilize business departments to think deeply about the use of AI at the heart of their activity and their strategy – we will see that Generative AI can be an accelerator to get there – however, this will not be enough; we must also use this exercise to capture the best ideas and identify the leaders capable of bringing them to light, understand AI in its complexity and carry out these initiatives while remaining focused on the business impact, which brings us to the second issue.
Remove human, organizational and institutional obstacles, whether internal or external, it is simply recognizing that we, our organizations and our ecosystems are not yet ready for the adoption of AI on a large scale and that it is necessary to anticipate the various obstacles to its deployment in order to better remove them or , on occasion, bypass them. We have an emotional relationship with artificial intelligence: with each new innovation, it materializes our fears about the limits of our capabilities, even about our humanity, as well as our fears going as far as the apocalyptic scenarios that some attribute to it. It is true that AI brings its share of slip-ups, such as, for example, Tay, Microsoft’s short-lived Chatbot and its racist remarks in 2016 which led to its immediate withdrawal from the market; a Chatbot that is strangely reminiscent of the hallucinations of ChatGPT and other Large Language Models (LLM). However, without the will and leadership of teams, AI will not be able to contribute to business results. Humans are the starting point for the adoption of AI and for them to be able to contribute to it as best as possible, many blockages and constraints will have to be removed both within the company, as well as in its markets or again with its ecosystem.
Obviously, these issues must be nuanced between different sectors and, perhaps even more, between different companies in the same sector. For example, Ping An, the Chinese financial services giant, stands out widely in the insurance sector in which many players are barely thinking about the subject. Let us also note regional differences associated with market size, regulations and culture of innovation which explain, for reasons specific to each, a more advanced maturity in Anglo-Saxon countries… and China.
Before delving into these issues, we must validate the relevance of William Gibson’s vision for AI. In other words, is the AI revolution real or utopian?
Patrick Darmon is a partner at Fizz venturecabinet de council and Data & IA.
1711238770
#put #business #model