🧠 Grok Connected to Chip with Live Neurons

AI blogger Garret "Artist" has introduced BioLLM, the first chatbot powered by real human neurons.

It uses a regular Grok through the API as the base. Every request is turned into electrical pulses that are sent to the CL1 bioprocessor from Cortical Labs. On this chip, 200,000 human neurons have been grown. They form a "living neural network": they react to external signals, adapt to previous inputs, and actually learn.

The neurons' reactions are visualized and converted into metrics that are fed directly into Grok's system prompt. The cells' activity, how well they're connected right now, and of course, the original request all play a part. In every next reply, Grok takes the current state of the biosystem into account.

🔎 Of course, BioLLM shouldn't be taken literally as a "living neural network." Still, the activity on the chip isn't just random noise—it really shows a biological system at work and how it adapts to the conversation.

@hiaimediaen