autocomplete
The task of predicting and filling in the next piece of text given preceding text; the training objective of large language models.
As defined in Erik Hoel on the Threat to Humanity from AI
Where it comes up · 4
- Erik Hoel on the Threat to Humanity from AI3 claims
- Kevin Kelly on Advice, AI, and Technology1 claim
- Is AI's "intelligence" an illusion? | GZERO World with Ian Bremmer1 claim
- Scott Wu - Building Cognition - [Invest Like the Best, EP.402]1 claim
Most important claims · 6 of 6Search all claims for “autocomplete” →
- ChatGPT's autocomplete mechanism shows that writing a sentence or essay is a more primitive and elemental thing than we thought—not as big an achievement as imagined—just as chess turned out to be more mechanical than expected.Kevin Kelly on Advice, AI, and Technology
- Large language models analyze the statistical relationship between words in their training data rather than analyzing concepts, ideas, or entities in the world, functioning essentially as 'autocomplete on steroids' that predict the most likely next word based on learned patterns.Is AI's "intelligence" an illusion? | GZERO World with Ian Bremmer
- Dismissing large language models as 'just autocomplete' is a mistake; the danger is not the text-completion task itself but the trillion-parameter neural network trained to do it, which is a mathematical black box capable of far more than autocomplete, just as humans were optimized to spread genes yet do dangerous things on the side.Erik Hoel on the Threat to Humanity from AI
- Humans also operate by a kind of effective autocompletion and our brains are likewise black boxes—Beethoven always knew what note should come next, much as ChatGPT does—so 'it's just autocomplete' doesn't prove the AI isn't smart, but neither does the brain's similarity prove the AI is a sentient brain.Erik Hoel on the Threat to Humanity from AI
- GitHub Copilot-style autocomplete saves an engineer the time of typing a predicted line, but it does not save the time of deciding what the correct line should be, because the human still judges correctness.Scott Wu - Building Cognition - [Invest Like the Best, EP.402]
- ChatGPT generates text by predicting the next word not necessarily as the single most common word that follows (which would read flat), but sometimes a surprising word, which gives the feeling that it is doing something thoughtful—a slightly smarter version of Gmail's suggested reply completions.Erik Hoel on the Threat to Humanity from AI