• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
AI
Reaction

AI ‘wrappers’ and the gap between capability and everyday use

A user points to Instinct and Muse as examples of apps that make OpenClaw capabilities more accessible to people without developer skills.

EmadEM
2 Sources, 11d ago, first seen 11d ago

TLDR

One post describes a “diffusion gap” between what AI can do and how most people use it. The writer argues that Instinct and Muse made capabilities already available in OpenClaw easier to use, and predicts more such apps could expand what consumers can do even if AI model capabilities stall.

Combined views

960

2 Sources, first seen 11d ago

10 likes2 comments

Combined views

960

2 Sources, first seen 11d ago

10 likes2 comments

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Featured Source

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Related

AI Founders With Revenue Shift Into Neoclouds

Posts note AI founders with revenue turning into neocloud providers.

Brad Lightcap Leaves OpenAI to Start New Venture

Colleagues reply after he posted plans to start something new.

Generated editorial image
Are there enough qualified auditors for frontier superintelligence models?

A user questions whether even a few hundred people could independently audit and evaluate such models, and whether they would need enormous computing power to do the job properly.

2 Sources

Jamin Ball@jaminballThe Diffusion Gap has been talked about for quite some time (shoutout @gradypb from his talk he shared last week). The general idea is a huge opportunity exists between AI capabilities and how the masses are using AI. With the stated implication being the vast majority of people aren't using AI to its fullest capability imo nothing demonstrates this better than Instinct / Muse. These capabilities existed with OpenClaw, and have exited for a while. BUT - it was hard to use. You needed to be a developer to make OpenClaw useful. The diffusion was low. Some call Instinct / Muse just an "OpenClaw wrapper." But that's exactly the point! Instinct / Muse added a lawyer on top of OpenClaw that helped close the Diffusion Gap by making the technology actually accessible to the masses (and fwiw these "wrappers" are oftentimes extremely complex) My expectation is that the next few years will see many "wrappers" emerge that help solve the diffusion gap. Model capability could completely stall and we'd still see an explosion of capability at the application / consumer layer11d
Emad@EMostaque@jaminball @gradypb I mean a good part of this must be that the models until just now were simply not competent enough11d
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    Jamin BallOpenClawEmad Mostaque

    2 Sources

    Jamin Ball@jaminballThe Diffusion Gap has been talked about for quite some time (shoutout @gradypb from his talk he shared last week). The general idea is a huge opportunity exists between AI capabilities and how the masses are using AI. With the stated implication being the vast majority of people aren't using AI to its fullest capability imo nothing demonstrates this better than Instinct / Muse. These capabilities existed with OpenClaw, and have exited for a while. BUT - it was hard to use. You needed to be a developer to make OpenClaw useful. The diffusion was low. Some call Instinct / Muse just an "OpenClaw wrapper." But that's exactly the point! Instinct / Muse added a lawyer on top of OpenClaw that helped close the Diffusion Gap by making the technology actually accessible to the masses (and fwiw these "wrappers" are oftentimes extremely complex) My expectation is that the next few years will see many "wrappers" emerge that help solve the diffusion gap. Model capability could completely stall and we'd still see an explosion of capability at the application / consumer layer11d
    Emad@EMostaque@jaminball @gradypb I mean a good part of this must be that the models until just now were simply not competent enough11d
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet