Opinions expressed by Digital Journal contributors are their own.
What if the next leap in artificial intelligence doesn’t come from scale at all? What if it comes from habit and structure? From teaching machines to build habits the way humans do?
According to philosopher Josh Adler, cognitive scientist Fei-Fei Li and DeepMind founder Demis Hassabis true intelligence, whether human or artificial, doesn’t come from raw computation but from the rhythms that sustain it.
AI needs behavior and not code
While AI has outpaced our expectations in precision and speed, it still lacks rhythm, consistency, reflection and prioritization. Fei-Fei Li, who pioneered modern computer vision, once declared, “If our era is the next Industrial Revolution, as many claim, AI is surely one of its driving forces.” Yet even she warns that technological growth without ethical and behavioral grounding will definitely create imbalance.
Li says that AI basically needs a moral and contextual framework, similar to how we need habits to translate our intention into action.
Adler puts it simply: “We don’t need machines that think faster. We need machines that remember better. Habit is memory with direction. It’s data that’s learned how to repeat itself in useful ways.”
Fei-Fei Li’s empathy machine
As per Fei-Fei Li’s research, AI models that trained on human-curated data successfully inherit our cognitive biases.
Habit architecture, as per Li, is how we teach machines to notice those patterns and correct them. In us, habits are what make up our character. In AI, they could become the baseline of ethics. This concept in itself is groundbreaking. Josh Adler extends that analogy: “Habits are compression algorithms for consciousness. If we can translate that efficiency into code, we get systems that learn like us iteratively, emotionally, efficiently, but without the chaos.”
Building machines that know when to take a pause
An interesting point emerges from all three thinkers, which is that intelligence requires hesitation. Systems that take a pause, recalibrate, and redirect themselves, behave more intelligently than those that run endlessly forward.
Adler adds: “Every good habit starts with a pause. Machines need that too. A point where they can ask, ‘Should I?’ not just ‘Can I?’”
The science behind the concept
As per the specialists, the brain doesn’t forcefully store data. It alters it, it builds heuristics, routines and habits. Adler believes that insight could define the next generation of AI design.“Right now,” he said, “we treat AI like a calculator. We feed it problems, it spits out the answers. But as humans we don’t just solve. We build patterns that let us predict what comes next.” He calls this habit architecture. It’s the scaffolding that holds intelligence in place long enough for it to grow.
A future of restrained intelligence
Entrepreneur Josh Adler takes this as a cue to make a call for restrained intelligence. “The problem with code isn’t that it’s wrong,” he said. “It’s that it’s one-dimensional. Habit is multi-dimensional. It includes context, memory and emotion. That’s where real intelligence lives.” He believes the next generation of developers won’t just be coders, they’ll be behavioral architects. They will be engineers of rhythm and designers of reflection. The goal is to make the code multi-dimensional.
AI that forms habits
Certainly, the race for more parameters, more data, more computers has reached its limits. As these three thinkers suggest, the future is all about alignment.
Thus, artificial intelligence doesn’t need more code, it needs cycles of learning, rest and recalibration. Just as we depend on physical resilience and mental discipline, machine intelligence may depend on algorithmic restraint.
