去中心化的AI：使用户恢复自主权.pdf

1、Decentralized AIBringing user self-sovereignty backCurrent AI landscapeCentralized vs Decentralized AIASI vs HSIUser self-sovereignty and cognitive libertyDecentralized AI implementationOutlineAI has been around since the 90s in more narrow scopesNow reaching the tipping point of widespread adoption

2、Acceleration in compute power availability and concentration of AI capabilities in the hands of few entitiesDecentralized alternatives still in the early stagesUsers lack the understanding around their data and privacyWhat is the path forward?Current AI LandscapeDominated by large tech companies suc

3、h as Google,OpenAI,Anthropic,Meta,etc.Data,compute power,development talent in the hands of fewCentralized AIAdvantagesEconomies of scaleAccess to dataRapid developmentDisadvantagesData privacy concernsMonopoly over AIRisk of biasRemoving single point of failure and adding transparency/verifiability

4、Users are in control of their dataDecentralized AIAdvantagesControl of dataEnhanced data privacyLess potential for biasDisadvantagesLower performanceSlower coordinationLack of standardizationData ownership with consent and individual controlTransparency in use of data and AI decision makingPersonal

5、data is valuable-user inclusion in the AI economyUser Self-SovereigntyCe-AIData breachesSurveillance and profilingLack of transparencyDe-AIUser controlled data sharingTransparent decision-makingDistributed governanceASI becoming a reality faster than expectedFundamental changes to social and economi

6、c structuresHow to compete with machines?HSI-Augmenting human intelligence with AIBrain-computer interfaces(BCI)AI-enhanced decision makingLearning and skill enhancementASI vs HSIWho has access to your thoughts?Privacy and freedom of thoughtThe right to control your own cognitive