On-chain AI and the future of personal privacy: Jesse Glass talks identity, AGI, and Decide AI’s 2025 roadmap News Desk · 12 hours ago
With its concentrate on privacy-preserving identity confirmation, Decide AI might redefine information defense in the Web3 period.
Dec. 20, 2024 at 2:00 pm UTC
In a current episode of the SlateCastCryptoSlate’s Editor-in-Chief Liam “Akiba” Wright and CEO Nate Whitehill took a seat with Jesse Glass, Lead AI Researcher at Decide AI. The discussion looked into Decide AI’s groundbreaking efforts to combine expert system (AI) with blockchain innovation, concentrating on privacy-preserving identity confirmation, the function of decentralized AI ahead of time AGI (synthetic basic intelligence), and the future of information ownership.
Changing Identity Verification
Jesse Glass opened the conversation by discussing Decide AI’s core objective: to produce AI-driven applications that confirm human identity without keeping delicate information on-chain. This technique makes it possible for “civil resistance” and makes sure greater information quality while keeping user personal privacy.
“Decide AI is a set of AI applications and combinations for LLMs and identity. For ID, we utilize AI to validate that you’re a human without saving that information on-chain,” stated Glass.
The system incorporates with both blockchain and Web2 applications, using an unique option to information defense. With increasing analysis on how individual information is managed by central entities, Decide AI’s method might function as a plan for privacy-first confirmation procedures.
AI on the Blockchain: Challenges and Possibilities
Releasing AI on-chain is a considerable difficulty, however one that Decide AI is actively attending to. Glass shared insights from his experience releasing GPT-2 on the Internet Computer (ICP) blockchain. He highlighted the technical obstacles presented by blockchain’s computational restraints, keeping in mind that standard AI workflows depend on off-chain calculate resources offered by big corporations.
“When I explore releasing GPT-2 on the Internet Computer, it exposed how ruined we’ve ended up being contracting out calculate to megacorporations,” stated Glass.
“On-chain AI provides difficulties like quantization, CPU cache vs. directions, and WASM collection, which needs enhanced information structures.”
Glass’s viewpoint highlights the shift needed to make AI more decentralized. While standard AI designs are supported by effective central facilities, on-chain AI needs more effective calculation and unique optimization approaches. This paradigm shift is not simply technical however philosophical, embodying the wider Web3 values of decentralization.
Open Source vs. Proprietary AI Development
The discussion moved to the future of AGI and whether open-source efforts might surpass exclusive AI advancement. Glass argued that the open-source neighborhood has a much better possibility of accomplishing AGI due to its collective nature and access to varied datasets.
“I’m not persuaded exclusive business will attain AGI. They’re too concentrated on revenues. The open-source neighborhood has prospective since AGI needs big datasets, cooperation, and contributions from numerous groups,” Glass mentioned.
He stressed that AGI advancement may look like scholastic research study– open, collective, and driven by shared objectives– instead of a closed, profit-driven race by tech giants.
