A reasoning gym that adapts to each user, where every question is generated anew, each one centered on a single distinction, and evaluated on the standard intelligence scale.
An application that hones the distinctions essential for sound reasoning could genuinely enhance your cognitive abilities. The underlying mathematics is transparent, and if the entire concept is flawed, the scoreboard will reveal it before anyone else.— Conner Lambden
IntelligenceMax, a reasoning application that generates each question uniquely for every user and evaluates responses on the standard intelligence scale, has been officially released today.
For years, brain-training applications have focused on honing reflexes. IntelligenceMax presents a single, precise multiple-choice question at a time, each constructed around a key distinction: options that appear similar but diverge on the single detail crucial for optimal truth-seeking. That detail might be an opportunity most people miss, a sunk cost that should be disregarded, or an unspoken incentive. Every incorrect answer is rooted in one of these blind spots, and the accompanying explanation identifies it in straightforward terms. The mistake itself becomes the learning opportunity.
No question bank exists, meaning nothing can be memorized and no pattern can be exploited. Each question is generated uniquely for an individual, based on that person’s previous responses, and the specific model responsible for each question is identified on the card. This allows the AI to design questions that are slightly beyond the user’s current comprehension level, thereby pushing the boundaries of their understanding. The question pools are refreshed weekly to align with the newest model releases, ensuring that the authors are always the most current top-tier options. God mode utilizes the strongest set of models, including GPT-5.6 Sol, Fable 5, Gemini 3.1 Pro, and Kimi K3. Plus employs a mid-tier group featuring GPT-5.6 Terra and DeepSeek. Free uses the fast set, GPT-5.6 Luna and Gemini Flash Lite. AI questions tailored specifically for you are the core of the application, and the challenge is structured so that a learned pattern never works more than once.
IntelligenceMax addresses the fundamental issue that the brain-game industry created: users improved at the games themselves, not at actual thinking. In 2016, the FTC reached a $2 million settlement with Lumosity regarding allegations that its games could enhance memory, focus, and academic or professional performance (Federal Trade Commission v. Lumos Labs Inc., Case No. 3:16-cv-00001, N.D. Cal.). Subsequent research confirmed the same conclusion
