In an earlier post I sketched what I dubbed the Buddha equation, a way of interpreting expected free energy (EFE) in active inference as a balance between two forces:
$$\mathcal G[q; \pi] = D_{KL}[q(o \mid \pi) \mid \mid \tilde p(o)] + \mathbb E_{q(s \mid \pi)} \mathbb H[p(o \mid s)]$$
- The first term (the KL divergence) measures the distance between predicted observations and preferred observations. This I likened to craving – the drive to make reality conform to a set of desires or priors.
- The second term (expected entropy) measures the uncertainty in predicted observations. This I compared to mindfulness or openness – an attentional stance that welcomes new precise data, even if it is not preferred.
In optimal active inference, the two terms are balanced. But human cognition, tuned by evolution, is likely often biased toward the first: the craving-driven imperative to survive, reproduce, and seek advantage. This makes sense from a genetic fitness perspective, but it also locks us into a search for instant gratification and a state of chronic dissatisfaction, since preferences are never stably met.
Enlightenment as recalibration
Buddhist practice can be seen as a re-weighting of the terms in the human (non-optimal) implementation of the Buddha equation:
$$\mathcal G[q; \pi] = \lambda_{\text{crave}} D_{KL}[q(o \mid \pi) \mid \mid \tilde p(o)] + \lambda_{\text{mind}} \mathbb E_{q(s \mid \pi)} \mathbb H[p(o \mid s)]$$
Here, \(\lambda_{\text{crave}}\) and \(\lambda_{\text{mind}}\) are weights reflecting the relative precision humans assign to preference-discrepancy vs. mindfulness.
- In the ordinary mind, \(\lambda_{\text{crave}} \gg \lambda_{\text{mind}}\). We give too much weight to the ego’s demands and treat any deviation as intolerable.
- In the awakened mind, \(\lambda_{\text{crave}}\) is reduced, softening the grip of craving, promoting equanimity, while \( \lambda_{\text{mind}}\) gains more influence, encouraging mindfulness, openness, and exploration.
Why this matters
Evolutionary rationale: High craving made sense in the harsh ancestral environment – it drove vigilance and competitiveness. But it also made us half-neurotic, forever restless. A too large \(\lambda_{\text{crave}}\) leads to overconsumption, anxiety, and eventually to poor mental health.
Cognitive rationale: Craving is prediction error given excessive precision. By down-weighting that precision, suffering diminishes while survival behaviors remain adaptive enough for today’s world.
Buddhist rationale: The cessation of suffering comes from attenuating craving (taṇhā) and realizing non-self (anattā). This is the 8-fold path in Buddhism. In AIF terms, this is recalibrating the weighting of terms in the Buddha equation.
Not nihilism, but efficiency
\(\lambda_{\text{crave}}\) should not be set to zero. We should only recalibrate the relative importance given to ego-driven preferences. The result would (and should) not be indifference to life, but a life lived with less friction, less manufactured suffering, and more equanimity.
In short: Enlightenment can be framed as a meta-optimization of the Buddha equation – a recalibration of the balance between craving and mindfulness, achieved by reducing the excessive weight evolution gave to the former.