TypeSafe AI, which is developing a model that outputs numerical responses with probability estimates to help businesses assess reliability, raised a $40M seed
Former OpenAI researcher Diogo Almeida says most intelligence is going to live inside software. But to automate any real work, it has to be correct.
TypeSafe AI has $40 million to develop and commercialize a model architecture aimed at giving business users probability estimates alongside numerical answers.
Enterprise teams seeking to automate bounded decisions gain a potential alternative to relying on unqualified model outputs, with reliability expressed as an input to their own acceptance thresholds.
Second-order effects
AI assurance vendors, including Artificial Intelligence Underwriting Company, have a more concrete output to assess if probability estimates can be tested against real-world accuracy.
Model providers and enterprise buyers face greater pressure to demonstrate calibrated reliability for production tasks, not just produce plausible answers.
Third-order effects
If such probability estimates prove usable in production, operational AI assurance may shift from broad claims of model quality toward measurable, task-level reliability controls.
The AI stack would increasingly differentiate between systems built for open-ended generation and systems optimized for constrained decisions that businesses can monitor and govern.
The trend: Enterprise AI is moving toward measurable reliability at the task level, linking model outputs to the assurance requirements for automated work.
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I've spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x [vid…
don't get one-shotted by this it's not a general language model and can't generate free form text it's probably a specialized diffusion model and it can only output a few different primitives and requires definitions of the output format
This is a big deal. Right now, people use LLMs a lot as classifiers in production systems. But they're slow, expensive, and hallucinate. This model is supposed to be >100x faster, at $42/BILLION tokens, and generates not only type-safe structured output but also calibrated confid…
diogo is an immensely creative guy and is working on really different types of models with the principle of building composable, programmable AI systems from layers of small inferences. it's a weird and ambitious idea thats worth tinkering with
we almost never test new foundation models but we've been testing this for ~a week @every and it's pretty wild. the kind of things that will be obviously indispensible in 6-12 months it doesn't produce words as output, it produces probabilities. so it can efficiently act as a jud…
This is incredibly cool. Completely new form of AI models—output tokens are so cheap to meter, they're literally free. The deflation of intelligence continues. 👇