Machines cannot fully replicate human intelligence in complex settings
Why machines will still not rule the world
Artificial Intelligence
Summary
The authors argue that artificial general intelligence—machines with human-like intelligence—is mathematically impossible. This is because human intelligence involves complicated systems and behaviors that cannot be fully captured by computer models. They also critique current arguments that suggest machine intelligence can rise by pointing out problems with theoretical assumptions and biases in testing. Their main point is that machines will not achieve true cognitive abilities in real-world complex environments.
What this means in practice
- •For machine learning engineers: Avoid overestimating AI capabilities when designing systems for unpredictable real-world tasks by considering limitations discussed in this paper.
- •For ai ethics teams: Use the argument that AI cannot fully replicate human cognition to inform policy and ethical guidelines around AI deployment.
A position paper. It proposes an approach and reports no results.
Authors
Jobst Landgrebe, Barry Smith
Abstract
In our book Why machines will never rule the world [13, 14] we argue that arti- ficial general intelligence is mathematically impossible. This is because the human beings and the processes which exhibit intelligence are complex systems whose be- haviour cannot be captured by the kinds of models that we can generate with or without computers. Proponents of contemporary machine intelligence respond with two lines of argument: a theoretical one, grounded in the universal approximation theorems for neural networks and the Church-Turing-Deutsch principle; and an em- pirical one, grounded in rapidly rising scores on standardized benchmarks. In this communication we examine and reject both responses. First, we show serious issues in the physicalist counter-argument based on the Church-Turing-Deutsch principle. Second, we review recent evidence to the effect that prominent benchmarks are compromised by training-data contamination, flawed test construction, and strate- gic optimization. Our central argument remains: That models required to perform cognitive behaviour in open-ended, thermodynamically complex and non-ergodic environments are not and will not become achievable.