“Block, the fintech company led by Jack Dorsey, has introduced an AI model called Dessa, trained exclusively on human conversational data rather than broad internet text. The model aims to produce more naturalistic dialogue, though early shared examples show interactions that can feel awkward and uncanny. This approach signals a growing interest in conversation-first AI development as an alternative to general-purpose large language models.”
Key Takeaways
- Block's new AI model, Dessa, was trained from scratch using only human conversation data, not general web text.
- Dessa's development is led by Ragavan Thurairatnam, who shared early human-AI chat examples publicly.
- Some published conversations appear awkward and uncanny, suggesting the model is still in early or experimental stages.
Jack Dorsey's Block unveils Dessa, an AI model built entirely from human conversation data.
trending_upWhy It Matters
Training an AI model exclusively on human conversation rather than broad internet corpora is a notable methodological departure from models like GPT-4 or Claude, which learn from vast mixed datasets. If successful, this approach could produce AI that feels more socially attuned, but the awkward early outputs suggest significant challenges remain in making purely conversation-trained models reliably coherent. For the AI industry, this is worth watching as a potential signal that specialised training data strategies may emerge as a competitive differentiator. End users and developers should pay attention to whether Block eventually opens Dessa to third parties or integrates it into its existing products like Cash App or Square.
FAQ
What makes Dessa different from other AI models like ChatGPT?
Unlike most large language models trained on broad datasets including web pages, books, and code, Dessa was trained exclusively on human conversational data. This specialised approach aims to make the model more naturally conversational, though it may limit its general knowledge and capabilities.
Why do some of Dessa's conversations look awkward or uncanny?
Training solely on human conversation means the model learns social and dialogic patterns but may lack the broad contextual grounding that mixed-data models have. This can produce responses that feel tonally off or socially stilted, especially in unfamiliar conversational contexts.
Is Dessa publicly available to use?
Based on the announcement, Block has only shared example conversations to demonstrate the model's capabilities, and there is no confirmed public release or API access at this time. Block has not yet announced plans for broader availability.
