New Frontier AI Model 'Jev' Impresses in Email Classification Test
Key Info
A newly released frontier AI model called Jev, trained with a new RLCD approach, was put to the test by Ryan Vogel, who says he tested it on 1,500 of his own emails and was "blown away" by its classification ability. The model is claimed to be 20-200x faster and 40-400x cheaper, with output tokens free.
Highlights
- Jev is described as a frontier AI model optimized for decision-making, built by a team that says it spent two years in stealth developing RLCD training.
- A practical test on 1,500 personal emails reportedly showed strong email classification performance.
- The project claims the shortest path to an AI-based economic revolution, though those claims are not independently verified.