Why does AI seem to be accelerating?
Two years ago, impressive new developments appeared each month. Now, it's every week.
All these news items come from one place: Silicon Valley. To see the future, you have to go to its heart, San Francisco.
That's what Side School did by sending Ben, their self-appointed explorer, with two questions to investigate: “What does the future of AI look like? And how do we adapt ourselves, our businesses, and France?” Here are his impressions:
1/ Everything stems from OpenAI
The creators of ChatGPT lead the way. Their APIs allow thousands of entrepreneurs to launch profitable businesses in turn (see below).
OpenAI's offices have shifted the historical core of Silicon Valley (Apple, Google, Stanford) towards San Francisco.
The rest of the world becomes peripheral.
2/ The new startup stars turning heads
Cursor, Lovable, Mercor: the promises of dot-com startups are now feasible. And the proof is verifiable: the best startups fund their growth with recurring revenue rather than fundraising. Cursor has 40 employees who have reached $100M ARR in 1 year: this growth becomes the new goal for the most ambitious entrepreneurs.
Ben during a hackathon organized by Anthropic (after the release of Claude 4)
3/ AI-native companies
Asking each employee to automate their own job with AI… A norm, already, in San Francisco?
In discussions with researchers at OpenAI and Anthropic, who are presumably at the forefront of progress, it's currently more of a geek dream than a reality.
But everyone is preparing to evolve in an “AI-native” company, integrating AI into every process and thought.
4/ Singularity: the fear we now take seriously
If we create an AI capable of improving itself, we potentially enter an “unchecked technological growth” spiral that's beyond our control. This acceleration phenomenon is called Singularity. A nerds' outlandish vision?
However, it is taken seriously here in San Francisco. The founder of Google, believes that we are already well into Singularity.
“AI 2027,” now the emblematic article of this vision, imagines a future where humans lose control of their destiny due to OpenAI's exponential growth.
Being in San Francisco is to be caught up in this fever.
But what is the likelihood of this scenario occurring?
5/ Choosing your side
Then, one must choose their side: to accelerate or to restrain.
The accelerators (who add “acc/” to their bios on 𝕏) capitalists advocate for the benefits of creating superintelligence and defending American supremacy against the Chinese. On the opposite, the doomer camp is rapidly growing, with political and lobbying actions.
Ben's Conclusions
The fervor of San Francisco will undeniably influence the rest of the world. The real risk is not that AI surpasses us (which it already does in some ways), but that we completely lose control over it. And as you might have guessed, it's therefore urgent to train ourselves to master prompting, agents, and the ethical limits of AI.
Auteur :
Biographie
Directeur Associé chez Side School. Ben Issen était précédemment fondateur de Supercreative où il a créé plusieurs outils IA à destination des freelances.