AI has already achieved human-level capability in many specific domains (coding, protein folding, materials science, chip design), and the scaling laws suggest continued improvement is inevitable without fundamental breakthroughs needed, so AGI emergence is a matter of 'when' not 'if'