The Gentle Singularity (AI column)
A couple of months ago, Sam Altman wrote an essay.Soft singularityIn which he shared his vision of the “soft” technological revolution. (It's funny that the phrase "soft revolution" is an oxymoron.). Sam believes that humanity has already passed the point of no return in the development of artificial intelligence and gained a second cosmic speed. But Sam's concept differs from the conventional apocapolyptical scenarios described by the conditional Jeffrey Hinton. (about the position I've already taken told).
In short, Sam makes this prediction about the evolution of possibilities in the next three years.
- 2025 Year - Agents capable of performing real cognitive work (code-writing)They have already changed programming.
- 2026 Systems capable of finding new ideas and generating scientific discoveries
- 2027 Robots that can perform tasks in the real world
In sum, the key idea of this essay is that the singularity occurs gradually, without dramatic upheavals. Sam describes the process as a smooth curve where "miracles become routine and then bets on the table." People are surprisingly quick to adapt to new possibilities, from admiring the ability of AI to generate a paragraph to waiting for the creation of entire novels. But these technological opportunities do not come alone; they are followed by economic transformation. Sam predicts radical changes in the economic structure, where 2030In the years to come, intelligence and energy will become abundant resources, removing two fundamental constraints on human progress. The cost of intelligence will eventually approach the cost of electricity through automation. The social sphere will also pull up behind the economy, but unlike the negative scenarios, Sam paints a utopian picture of the future, where technology will free a person from routine work. He draws a historical parallel: a medieval farmer would consider modern occupations “fake,” just as in a thousand years our descendants will be amazed at our occupations.
All in all, Sam is a techno-optimist in this essay, but he touches on a number of very important issues. The problem of alignment, or rather the need to ensure that AI serves long-term human values. Distribution of access to AGI - it is important to ensure wide access to superintelligence, rather than concentrating it in one hand Social Contract – It’s important to collectively define the boundaries of AI
Well, I rather liked the essay, but I see a gap between Sam's utopia and the current tactical state of affairs, where Business Insider writes about the decline of mega-employers and the shortage of white-collar jobs. (I've been looking into it. here). As a result, this “soft singularity” concept faces the hard reality of corporate optimization. While the long-term outlook may be positive, the transition period creates serious social strains. I think the key question is how to make the transition to an AI economy fair and manageable, ensuring not only technological breakthroughs but also social stability.
#AI #Software #Economics #Engineering #ML #Management #Leadership