🔍 Read the full analysis: The Role Of 'System One' AI In Shaping Future Technology Trends on ThorstenMeyerAI.com
Get smart everyday buys delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
TL;DR
TypeSafe AI has launched Jev, a new ‘System One’ AI model designed for decision-making within software. It emphasizes speed, cost-efficiency, and structured decisions, challenging traditional large language models. This development could reshape how businesses adopt AI for automation.
TypeSafe AI announced the release of Jev, a new class of AI model designed explicitly for decision-making within software systems, on September 15, 2026. Unlike traditional large language models (LLMs), Jev produces structured, typed decisions with associated probabilities, aiming to streamline automation processes and reduce reliance on text generation. This shift marks a notable departure from the prevailing trend of deploying chatbots and conversational AI in enterprise applications, signaling a potential transformation in how AI is integrated into business workflows.
Jev is built around the concept of ‘System One’ thinking, inspired by psychologist Daniel Kahneman’s distinction between fast, intuitive judgments (System 1) and slow, deliberate reasoning (System 2). Developed by TypeSafe AI with $40 million in funding led by DCVC, Jev is designed to handle structured questions—such as choices, scores, and yes/no decisions—and return typed answers with calibrated probabilities. This contrasts sharply with traditional LLMs, which generate free-form text that often requires parsing or human interpretation.
TypeSafe claims Jev can deliver responses in 70 to 500 milliseconds at a cost of approximately $0.042 per million input tokens, asserting that it is hundreds of times faster and cheaper than existing models. The company emphasizes that Jev’s outputs are schema-compliant, eliminating errors caused by formatting issues or hallucinations common in LLMs. However, accuracy benchmarks show Jev achieving around 62.6% agreement with reference models on certain tasks, indicating that precision depends heavily on how questions are structured and the quality of training.
Jev vs. LLMs: who should make the call?
Jev, from TypeSafe AI, is a “System One” model. It doesn’t write text. It returns a typed decision with a confidence score that your software can act on directly.
Same support ticket, two kinds of answer
“This ticket appears most likely related to billing, although it could also concern account settings or a recent plan change. I would suggest reviewing the invoice history before…”
A person reads it, or code has to parse the prose.
team: "billing"Software reads it and acts. Nothing to parse.
How they differ
| LLM | Jev | |
|---|---|---|
| Output | Text written for people | A choice, a score or a yes/no probability |
| Speed | Seconds per call | 70–500 ms* |
| Price | Input and (pricier) output tokens | $0.042 per million input tokens, output free* |
| Knows when it’s unsure | Often sounds confident when wrong | Confidence score on every answer |
| Explains its answer | Yes | No, which matters for audits |
| Best at | Reasoning, writing, open questions | Routing, tagging, scoring, duplicate checks |
* Vendor-reported. TypeSafe also claims up to 194× faster and 445× cheaper on its own selected workflows.
Accuracy is something you build
Jev is far cheaper and faster, but not more accurate than frontier models. How you phrase the question matters a lot.
TypeSafe’s benchmark scores agreement with two frontier models rather than verified ground truth. The five-question result used weights fitted on 1,000 labelled examples.
The real idea: a confidence dial you control
“duplicate listing”, confidence 0.62
Raise the threshold for fewer mistakes and more manual review. Lower it for more automation and more risk.
Only use Jev when all four hold
Good fits
- Routing tens of thousands of support tickets a day
- Flagging duplicate listings in a product catalogue
- Replacing a keyword filter that mis-tags half its matches
Poor fits
- Drafting customer emails or release notes
- Reviewing a few high-stakes contracts a month
- Anything that needs a written explanation
Implications for Enterprise AI and Automation
The introduction of Jev signals a potential paradigm shift in enterprise AI, moving away from text-based conversational models towards decision-focused automation. By producing typed, structured answers with confidence levels, Jev enables faster, more reliable integration into software workflows, reducing costs and increasing efficiency. This approach could expand the scope of automation, especially for routine decisions that currently require human oversight or complex parsing of language outputs.
Moreover, Jev’s emphasis on calibrated probabilities and schema compliance addresses some of the reliability issues plaguing traditional LLMs, such as overconfidence and hallucinations. For businesses, this means more predictable automation, less need for human intervention, and potentially lower operational risks. If widely adopted, Jev and similar ‘System One’ models could influence the development of future AI tools, prioritizing decision accuracy and speed over conversational complexity.
enterprise decision-making AI software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Evolution of AI Models and Industry Shifts
Over the past three years, the AI industry has seen a surge in large language models like GPT-4 and Claude, promising better reasoning, longer context, and improved code generation. These models primarily generate free-form text, which has been useful in chatbots, content creation, and customer service. However, limitations such as hallucinations, overconfidence, and high costs have prompted a reevaluation of AI strategies in enterprise settings.
TypeSafe’s Jev represents a departure from this trend, focusing on structured decision-making rather than text generation. Co-invented by Diogo Almeida, a key figure behind RLHF techniques used in ChatGPT, Jev leverages a new training approach called Reinforcement Learning for Calibrated Decisions (RLCD). This method aims to produce more reliable, predictable outputs tailored for automation rather than conversation, aligning with industry demands for efficiency and precision.
“Jev is designed to produce decisions, not words. It’s more like a function call within your software, and it’s faster, cheaper, and more reliable for automation.”
— Diogo Almeida, CTO of TypeSafe AI
structured decision automation tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unanswered Questions About Jev’s Effectiveness
While Jev shows promise in speed and cost, its accuracy benchmarks indicate room for improvement, with current agreement rates around 62.6% on some tasks. Its performance varies depending on input quality and question design, and independent evaluations have shown lower accuracy compared to some frontier models. Additionally, it remains unclear how well Jev will scale across diverse enterprise applications or handle complex, nuanced decisions in real-world settings.
Further testing and real-world deployment data are needed to fully assess its reliability, robustness, and potential limitations in different industries and use cases.
AI decision-making models for business
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for Adoption and Development
TypeSafe plans to continue refining Jev’s training methods, aiming to improve accuracy and expand its capabilities for more complex decision-making tasks. The company is also likely to seek broader industry partnerships to pilot Jev in various enterprise environments, testing its scalability and reliability in real-world scenarios.
Meanwhile, competitors and industry observers will monitor how Jev’s decision-focused approach influences AI development, possibly prompting a broader shift towards structured, automation-oriented models. The next few months will be critical in determining whether Jev can deliver on its promise of transforming enterprise AI workflows.
As an affiliate, we earn on qualifying purchases.
Key Questions
How does Jev differ from traditional large language models?
Jev produces structured, typed decisions with associated probabilities, rather than free-form text. It acts more like a function call within software, enabling faster, more reliable automation with schema compliance, unlike traditional LLMs that generate language and often require parsing or human interpretation.
What are the main advantages of Jev for businesses?
Jev offers significant speed and cost advantages, enabling decisions in milliseconds at a fraction of the typical model cost. Its structured outputs reduce errors caused by formatting issues or hallucinations, leading to more predictable and reliable automation of routine judgments.
What are the current limitations of Jev?
Current accuracy benchmarks suggest that Jev’s performance depends heavily on question design and training quality. Its agreement with reference models is around 62.6%, and independent tests show variability in confidence calibration. Further development is needed to improve its reliability across diverse tasks.
Will Jev replace traditional chatbots?
Jev is designed more for decision automation within software workflows rather than conversational AI. While it may reduce the need for some chatbot functions, its primary aim is to streamline decision-making processes rather than replace conversational interfaces entirely.
What is the future outlook for ‘System One’ models like Jev?
If successful, ‘System One’ models could reshape enterprise AI by prioritizing speed, reliability, and structured decision-making. They may complement or even replace some traditional LLM applications, especially where automation and efficiency are critical.
Source: ThorstenMeyerAI.com
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
