Loading activity…
@armenagha
Co-founder & CEO @perceptroninc; ex-RS FAIR/MSFT
AI
A user offers a brief positive take, calling the new DeepSeek model “very good.”
Technology
The company says the model improves benchmarks, task speed and output length over its predecessor.
AI
A user says a New York Times piece describes Anthropic quietly bringing in religious thinkers from around the world to help give its AI a moral compass.
AI
A post promotes demos it describes as showing models running on different embodiments.
AI
Perceptron says the model can fly drones, control four-legged robots and power smart glasses, with no platform-specific retraining.
AI
A user finds Astra noticeably better than Fable at theology, but says it still consistently makes significant errors on Christology and the Filioque.
AI
Researchers on X compare the new Explorative Modeling paper to earlier IMLE methods in algorithms, loss, and motivation.
