Presentation

After 75 Years of AI, Can Machines Think?

Tim Finin

April 1, 2024

3763180 bytes

PDF Document - Need a reader? Get one here

ai, chatgpt, generative-ai, llm

Mathematician Alan Turing proposed a simple test to answer the question 'Can machines think?' nearly 75 years ago. Today, the surprising abilities of generative AI systems like ChatGPT make many wonder if we can finally respond positively. Dr. Finin will briefly cover AI's history leading up to the recent development of systems using neural networks and large language models like ChatGPT and what to expect in the next few years. He'll touch on what current systems can and cannot do, the ways they can fail (e.g., hallucinations), the larger risks they may enable (e.g., spreading disinformation), and the societal changes they might induce (e.g., changing copyright and education practices). Dr. Finin will finish by outlining the prospects for the good things they will enable and how AI researchers hope to achieve them.


Presentation made online as part of the UMBC AOK library symposium series on AI, Privacy and Ethics in April 2024.

48 downloads

Public

OWL Tweet