Important: Translated automatically from Spanish by 🌐💬 Aphra 1.0.0
Without fanfare. Without making waves. Machines have conquered the penultimate bastion they needed to supplant humanity: reasoning. The great barrier that automation faced has been demolished. I’m referring, of course, to OpenAI’s announcement of o31.
It’s true that these models don’t reason as humans do; they only simulate doing so. But, setting our ego aside, the question is: How much does that difference really matter?
Many believed these models had hit a wall. It turns out we’ve jumped over it simply by doing something as straightforward (greatly simplified) as giving them “more time” (aka tokens2) to think based on the difficulty of our request. A stroke of luck for tech companies in the Artificial General Intelligence race. A problem for users, because AI processing costs are what scale and increase. Therefore, the upward trend is toward a more expensive and restricted “intelligence”. The computing time of these models is the new gold, and investments flow to whoever demonstrates they can generate more intelligence, faster and cheaper. Talent attracted by the number of GPUs an organization has. Never before in history could the highest bidder buy intelligence - even if artificial. Yet here we are. This is the situation that isn’t being discussed enough.
There’s still hope. Because the downward trend continues. Increasingly capable and cheaper open models. If nothing prevents it, this intelligence could be democratized, opening doors to new possibilities.
Let’s take scientific and technological advancement as an example. For a long time, most innovations haven’t come from individuals, but from teams of people. It’s necessary to combine disciplines and fairly specialized knowledge in them. But we have limited time to delve into different branches. That’s why teams achieve more. However, it also has its drawbacks: Different availability, funding searches, multiplied bureaucracy, learning to coordinate and understand each other… Perhaps this cheapening of intelligence will allow the resurgence of new Thomas Edisons or new Beulah Louise Henrys3. Individual people capable of using this technology to compensate for certain knowledge gaps and accelerate the discovery of new innovations.
Against this backdrop, we face a concerning issue on the other side of the coin. Due to the bombardment from various media outlets and clickbait headlines about AI’s potential social impact, many young people are experiencing an educational existential crisis4. I constantly see messages on various social networks expressing concern about how AI will affect their chosen field of study. For example, what’s the point of learning programming if we’re constantly told that machines will do this work sooner rather than later? Meanwhile, countless YouTube ads encourage these same young people to invest in dubious get-rich-quick schemes promising financial freedom by their mid-twenties.
Don’t fall for it. Your knowledge, your learning, has been, is, and will be your greatest source of wealth. We’re living in the moment in history when it’s clearest that the more you learn, the better. Critical thinking isn’t possible without knowledge. Learn about what you like and what you don’t like too. Never stop cultivating your digital skills, whether they’re directly related to your field (honestly, I can’t think of any field today where they’re not) or not. You can also use Artificial Intelligence as a springboard for your learning. The only thing that can keep you ahead of the curve is your ability to use new technologies as your tool. And for that, you need the best possible knowledge. In your chosen field, or better yet in several fields, and in technology. That’s why your response to “there’s no point in learning that” should be to learn even more intensely. It’s true that all jobs, or most of them, will be redefined in some way. The only way to survive that, the only thing you’ll be able to hold onto when that moment comes, will be your knowledge.
If we fear these advances will end in a machine revolution, it’s because we’ve been living in a cave5. Because that revolution started long ago. And all of us, in some way, are part of it.
Want to make a smart investment? Invest in yourself.
Happy New Year.
PS: No language models were harmed in the writing of this article… But one was used to generate the illustrative image.
-
o3 refers to GPT-4’s successor, announced by OpenAI for future release. ↩︎
-
In AI language models, tokens are units of text processing that allow for longer and more complex responses. ↩︎
-
Beulah Louise Henry, known as “Lady Edison,” was a prolific American inventor who held 49 patents, exemplifying successful individual innovation. ↩︎
-
A term describing the anxiety and uncertainty young people face about their educational and career choices in light of advancing AI technology. ↩︎
-
This reference to Plato’s Allegory of the Cave suggests a state of willing ignorance about technological change that has already occurred. ↩︎
