Those of us who look at AI from a historical perspective of the introduction of breakthrough technologies compare it with such revolutions as the diffusion of firearms through the late 15th and 16th centuries and the steam-powered industrial overturn in the second third of the 19th century. Despite its extraordinary outlook, AI has the same structure of incremental innovation, increasing its impact from step to step, with each of them proposing multiple options for development. Some of the steps, and often the most attractive of them, are dead ends, while the true game-changers remain concealed. The true game-changers are mainly not the products of the new technology but something else. Today, LLMs dominate the AI landscape, but they are only one among many AI types. Some alternatives look more promising while LLMs allegedly come to the limit of their efficiency. Artificial Intelligence, as an entity, looks like black gunpowder, a primitive mix of saltpetre, charcoal, and sulphur before its first “technological leap” of granulation in the mid-15th century. The massive data centres that LLMs use look like the first cumbersome wrought-iron guns on wooden logs before the bronze artillery on aiming carriages was introduced in the late 15th century. We can expect the similar “software” (types of AI) and “hardware” (kinds of its infrastructure) breakthroughs to come in a few years or decades. However, the first successes of firearms were grounded not in their technical sophistication but in the technique of their employment. It is the entity that we call a methodology today. The Frenchmen achieved the fantastic results during their invasion of Italy at the end of the 15th century due to the methodology of combining the guns in batteries and concentrating their fire against the small vulnerable parts of the walls to crumble down all construction. They stormed and sacked the towns like Pisa and castles like Naples, impregnable before. Then, a few years later, the Spaniards achieved the same fantastic results by assembling their arquebusiers in tight firing squads behind the moats and under the pikemen’s protection. The moats before the Spanish positions at Cerignola were stockpiled with the corpses of the “flowers of the French nobility”. From these first successes, the firearms almost did not improve up to the end of the 17th century, but the methodology of employment – the organisation of the troops and their tactics – changed fast and radically. The advantage belonged to the combatants that professed the better methodology and not to the ones deploying better arms. It is the base philosophy in which our methodology – Strategy by AI – is grounded. www.strategybyai.org It is combined with neuroscience’s belief in the crucial importance of the human brain’s asymmetry for cognition, reasoning, and action. We neither invent the new variations of AI (gunpowder) nor AI’s equipment (gun and handgun). We profess the methodology (organisation and tactic) to handle the AI device that is in abundant proposal (like first guns and handguns) – LLMs. Virtually speaking, we build a gun battery and dig a moat before the firing position for successful deployment of the existing AI. The results are sound in strategy analysis, making, and execution for business and political organisations. www.strategybyai.org