slashentry.com
AI lets anyone pass. Understanding has become optional.
Slash Entry is an education research project rethinking how we assess learning in a world where anyone can use AI. Built in the open by Dev and Design HQ.
/01
The problem
Exams were built on a simple assumption. What a candidate produces reflects what they know. AI has quietly broken that assumption.
A student can now score well above their real level of understanding. When the score stops reflecting the person, the result stops being useful. And a lot of important decisions are built on that result.
/02
Our stance
We are not trying to keep AI out of learning. AI is genuinely useful. It helps people learn faster, think through harder problems, and work in ways that are not going away. Banning it is neither possible nor wise.
The problem is not that people use AI. It is the growing gap between using AI and actually understanding. That gap is the whole of our work.
/03
What we are doing
Slash Entry researches new ways to assess understanding in an AI-augmented world. The goal is assessment where genuine understanding stays a requirement for success, not an optional advantage, even when a candidate has full access to AI.
This is applied research. What we learn is meant to be built and used, not only written up.
/04
RESEARCH ABSTRACT
Redesigning Learning and Assessment for an AI-Augmented World
AI is quietly breaking assessment. Exams, tests, and evaluations were designed for a world where a candidate's output reflected their own understanding, and that assumption no longer holds. With AI, candidates can now score well above their actual level of comprehension. This means test results no longer reliably show what someone actually knows. Schools, employers, and licensing bodies depend on those results to make real decisions, so when the results stop being trustworthy, so do the decisions: admissions, hiring, certification. But the answer isn't banning AI from assessment. AI has already proven genuinely useful in education, enhancing self-learning and sharpening thinking, and it's becoming a permanent part of how people learn and work. The real problem is the gap between AI use and demonstrated understanding, and that gap is what this project exists to close. We are researching assessment mechanisms that let candidates use AI freely while maintaining genuine understanding as a requirement for success rather than an optional advantage. This work is applied and product-facing: findings will directly shape the testing and evaluation tools Slash Entry builds alongside theoretical frameworks. Restructuring assessment now, at this early stage of generative AI, is crucial. This shift can't come after the fact; assessment systems have to evolve alongside AI so the disparity between the two never gets too wide. The further apart they drift, the more assessment measures a world that no longer exists, and the harder it becomes to close that gap later.
/05
EARLY / EXPLORATORY
Status
This is early work. We are mapping approaches and testing ideas, not reporting finished results. We are building in the open and sharing our thinking as it develops.
If you care about this problem, now is a good time to follow along.
/06
Why it matters
Assessment is quiet infrastructure. Most people never think about it, but it decides who gets into school, who gets hired, and who gets trusted to do serious work.
When the measure stops working, those decisions get made on bad information. Fixing this while generative AI is still young matters, because the longer assessment and AI drift apart, the harder the gap is to close.
/07
Who is behind it
Slash Entry is a research project by Dev and Design HQ, a company working at the intersection of design, engineering, and AI. We train people to build real products with AI, which is where we ran into this problem up close.
/08
Follow the research
We share what we find as we find it.
Get in touch →