Artificial Intelligence

AI tools are constantly evolving and vary in capability, ease of use, and cost. Some can be downloaded and others are product integrated (e.g. Copilot in Microsoft 365 apps). The free apps do not have the power of those requiring a subscription.

AI tools can support your research activities such as brainstorming  for research ideas, developing research questions, suggesting database search terms, deconstructing challenging concepts, simplifying complex language, interpreting data sets, or suggesting an essay structure.

• Artificial intelligence describes the broad concept of machines simulating human thinking using computer algorithms. It is machines programmed by humans. AI is not a search engine or an information repository. Term originates from the 1950s but new forms now emerging rapidly.

• Machine learning uses AI but goes beyond computer programming to autonomously use complex algorithms and statistical models to identify patterns within large data sets, and to evolve and adapt these algorithms through repeated experiences.
• Generative AI (e.g. ChatGPT, DALL-E) is trained using a Large Language Model which draws on expansive BIG data and functions with human-like responses. It generates text, images, and sound, based on its training data and user inputs. Outputs include letters, essays, poetry, stories, video, music, art, computer code. 
• Large Language Models generate outputs using neural networks that mimic brain interactivity by connecting and predicting letters and words. Data passes through many processing layers to enable interpretation and relationships. Words are converted into tokens which are mapped to numerical vectors.

This figure by Mukhamediev et al. (2021) maps out various branches of AI specialization.