High-Temperature Searches

I came across an interesting concept that touches on both neuroscience and artificial intelligence, recently.

My first contact with this concept was in a book about psychoactive substances, but the notion branches out to encompass a wide-variety of fields, depending on how you think about it.

The idea is that some thinking requires more brain power than other types of thinking. In particular, thinking our way through novel problems, or trying to come up with original solutions to existing problems—this taxes our brains more than coming up with rote or customary solutions that we can either recall or reference, rather than having to develop new ones out of nothing, from whole cloth.

What this means in practice is that, because our biologies do their best not to overtax us and aspire to save energy wherever possible, we tend to revert to norms, to the predictable, whenever possible. And we go the extra mile to try to cram unfamiliar problems, unfamiliar variables, into familiar boxes. This allows us to use less processing power, and to benefit from the knowledge we’ve accumulated over the course of our lives; both experiential knowledge, and learned, data-based knowledge.

Children, though, don’t have that brain-based library of experience and facts to pull from. So their brains tend to approach each new problem as, well, a new problem. Rather than contextualizing it as something familiar to make problem-solving more efficient, they start from scratch, extrapolate in wild directions, and approach it without any preconceived notions; no sense of what’s more probably correct and what’s more probably not correct.

That means, from the baby’s perspective, any path is a possible, legitimate path to take.

From the adult’s perspective, on the other hand, it’s supposed that some ideas will be more likely to work, while others are far less likely and therefore not even worth trying.

Children, then, seem to be more neurologically primed to explore, while adults may be more neurologically primed to exploit—to benefit in terms of energy and time saved from previously acquired experience and knowledge.

In some facets of the world of artificial intelligence, these dichotomous methods of exploration are referred to as high-temperature and low-temperature searches.

High-temperature searches require more energy, more processing power, and therefore cause the hardware hosting the AI to run at higher operating temperatures.

Low-temperature searches make use of existing knowledge and experience to come to conclusions more efficiently, cutting out the wild extrapolations and harebrained schemes that will almost certainly not work, to focus on the small collection of more-likely solutions.

There are pros and cons to both of these approaches, of course, but it’s thought—and this takes us back to the world of neuroscience—that the older we get, the more we tend to rely on low-temperature thinking and imagining. With age and experience, our thought-processes fall into grooves in the intellectual road that our prior trips have slowly weathered into the pathway of possibilities.

This means that we’re more likely to come up with probably-right answers more of the time, and faster, the older we get. But it also means, on average, that we’re less likely to make connections between seemingly unconnected things, because we don’t even try to do so. We’re less likely to come up with asymmetric approaches to solving problems, and we’re less likely to defy “the way things are” and “the way things work” as those norms are defined by our lifetime of experiences.

There’s almost certainly a social component to this, alongside this potential, neurological component. Who, after all, wants to be seen trying to drive a nail with a fish instead of a hammer? An adult human tends to know that the hammer will work better, but a baby doesn’t, and as a consequence, that baby may try a bunch of silly-seeming things, some few of which may end up being brilliant, while the majority of them only result in a very unhappy fish-hammer.

This was brought up in a book about psychedelics because it’s thought, by some, that these sorts of substances can help put us adults back in a high-temperature thinking space, where the spectrum of what’s possible and likely is more expansive, just like when we were kids. That sense of play, that sense of potential, increases, unhindered by the grooves our mind-tires have eroded over the years, giving us the ability to think more freely, more inefficiently, for a time.

This is primarily speculation at this point, but it’s an interesting thought. And although I don’t have any strong feelings against psychoactives, it’s worth considering how we might achieve the same—part of the time, at least—without needing any such substances to get us there; without relying on the external to expand our internal capabilities.

Might it be possible to liberate ourselves from grooved-thinking by simply achieving the proper mindset? Might it be possible to expand our sense of possibility by meditating or breathing the right way, or by having varied enough experiences that we sometimes bypass our sense of shame and concern over being seen as foolish? So that we might justify being more adventurous and “childish”?

It’s an idea that’s worthy of consideration, and a variable that may be worth working into our personal conceptions of self and possibility.

It’s relatively common to aspire to be more capable, I think, but less common to aspire to be less efficient and more meandering in how we problem-solve.

Periodic intellectual deviations, though, could prove to be valuable and enjoyable additions to our habitual thought-patterns.

If you enjoyed this essay, consider supporting my work by becoming a patron or purchasing one of my books.





Recent Posts

  • Tensions
  • Which Ends?
  • What’s My Thing?
  • Unlocked
  • Lustrous Tools