“AI: The only thing worse at remembering than me.”
What happened: Scientists developed a new test to evaluate AI chatbots' understanding of relationships. Claude 4.6 scored 73%, outpacing GPT-5.4 at 68%, but both struggled with memory and reasoning, especially in astronomy. The test highlights areas where AI needs improvement before we rely on it for tough decisions.
This Neural Toon is based on the story “SciReC: Diagnostic Evaluation of Multimodal, Multi-Turn Relational Reasoning with Adaptive Interaction”.
Read our summary of the story · original source
Panel 1 – Jordan: AI can ace any relationship test!
Panel 1 – Taylor: Even my ex couldn't do that.
Panel 2 – Jordan: Claude 4.6 scored 73%.
Panel 2 – Taylor: 73%? That's like a D- in relationships!
Panel 3 – Jordan: Now it knows your secrets!
Panel 3 – Taylor: Great! It’ll tell my mom I'm single.
Panel 4 – Jordan: Trust AI for complex questions.
Panel 4 – Taylor: Right, like trusting a cat with a laser pointer.
More Neural Toons · Neural Snaps · The Neural Feed
One email a day, no spam, unsubscribe anytime.