Generation AI - by Vincent Barat

Episode 3: It’s AI-okay, don’t you go all Terminator on me

By enabling our clients to plan their staffing and skills needs, we at Albert are also at the forefront of the transformation of companies through AI.

It is a subject that we have already covered extensively, notably in a publication in Mag RH dedicated to the Promethean myth of AI (in French, though, sorry). In the face of countless anxiety-inducing predictions announcing the massive disappearance of jobs as a result of artificial intelligence (AI), let us remember that these prophecies are not new. Already 10 years ago, similar studies were predicting an employment catastrophe caused by digitization. However, reality has shown that these predictions were greatly exaggerated: the number of jobs considered to be at high risk has only decreased slightly, and some have even increased in volume. You might say that this fear business is a living.

The truth is that nobody knows. Our clients, who are more like large groups that cannot be accused of not being far-sighted and proactive - they use Albert after all - are all testing hypotheses. But none of them are brimming with certainty. And even in scale-ups, some radical visions can be questioned. Klarna, for example, drastically reduced its workforce last summer, and notably replaced its customer support with AI. The feedback became so negative that they backtracked and ended up rehiring humans. According to the CEO, “nothing will be as valuable as people”. No shit, Sherlock...

Without betraying any secrets but to remain specific, there are two types of hypotheses under consideration:

• Productivity gains in all professions where there is “file management”. But no massive layoffs. We are generally talking about gains in the order of 5% to 10% in the most extreme scenarios. In regulated environments in particular, human controls are still required anyway.
• Changes in skills. Everyone will have to get on board. It's no different from the issue of digitalization. And it is not without danger: as with digitalization, some people will find it difficult or refuse to adapt and will be left behind. But unlike the wave of digitization, another danger will arise - apparently very accentuated in the working classes according to recent studies: a certain form of credulity and lack of critical thinking, which will lead users to take everything that comes out of Gen AI at face value.

On our micro-scale of startups, we see the same thing. Having personally developed technical “Proof of Concept” prototypes, I can testify that we are not about to eliminate the role of developer. I do have a technical background, however, and even without having been a developer, I have some knowledge of code, just as I have some knowledge of Latin: I can decipher and do a little versioning, but I would be incapable of writing The Very Hungry Caterpillar. Except that with 500,000 lines of code in Albert, it's more like War and Peace that our developers write every day. To quickly test an idea, the advances in AI are therefore formidable, but to scale up, respect good development practices, secure and make an application robust, that's another story.

The same is true for marketing content. While we can “streamline” the production of certain posts or certain SEO-oriented content, we cannot do without authentic production. All of these publications were produced without generative AI, because it will not be able to imitate us or retain the incisive edge found in “organic” content.

Finally, it is striking to note that there is still little real training in large companies or in the education system associated with the deployment of these new technologies. I am not talking here about generic inspirational speeches or vague awareness-raising. But real courses on how to tame and make the best use of these technologies, which are still frowned upon and prohibited in many contexts. Unfortunately, having been a little “slow on the uptake” is what has created our current situation: the disintegration of critical thinking online, and stupefaction through social networks. Let's try to do better.

To read or reread the previous episodes of this series of articles devoted to the “Generation AI”, click here.