Global | Large Language Models, with data from all and... for all

Large Language Models have revolutionized the Artificial Intelligence (AI) landscape since they were first developed in 2017 when a group of Google researchers wrote their now famous "Attention is all you need".

Key points

• Key points:
• Large Language Models have the ability to learn patterns and relationships between words and phrases to predict text. Their use has already gone farther, and the latest versions are multimodal, integrating text, images and audio.
• They are a very useful tool for work, with the potential to generate clear productivity gains in our daily activities.
• In fact, analyses of their use suggest that almost 70% of chat conversations and 40% of searches in tools developed with large language models are aimed at professional tasks.
• Recent empirical analyses, such as those by Brynjolfsson and colleagues at Stanford last year, reveal that, aided by generative AI, worker productivity increases by about 15%.
• The "cognitive automation" of repetitive tasks is already achieving significant productivity benefits; and as analysts and researchers use them in "open source" and we improve their efficient usage techniques, and their suitability, the gains from these models will be even greater.

Authors

Geographies

Topics

Tags