latest crypto ai stories, powered by the
@useDroyd feed
1)
@sama announced
@OpenAI raised $110 billion from Amazon, NVIDIA, and SoftBank. Amazon is investing $50 billion to co-develop a stateful runtime on Bedrock and supply 2GW of Trainium capacity.
2)
@AnthropicAI rejected a final offer from the US Pentagon for unrestricted military use of its AI. CEO Dario Amodei cited concerns over mass domestic surveillance and fully autonomous lethal weapons.
3)
@jack announced
@blocks is cutting 4,000 jobs, nearly 40% of its workforce, to restructure around AI-driven efficiency. The move signals a shift toward smaller, flatter teams powered by intelligence tools.
4)
@Google launched Nano Banana 2, a low-latency image generation model with 4K output and world-knowledge grounding. The model is rolling out across
@GeminiApp, Search, and developer tools via
@GoogleAIStudio.
5)
@Alchemy launched Crypto APIs for Agents, enabling autonomous AI agents to self-signup and pay for compute in
@USDC via x402. The system supports read/write actions across 100 blockchain networks.
6)
@0G_labs and Stanford blockchain veterans launched the $20M Apollo AI Accelerator. The program provides up to $2M in funding and
@googlecloud credits to scale revenue-generating decentralized AI applications.
7)
@perplexity_ai partnered with Samsung to integrate its AI as a system-level assistant on the upcoming Galaxy S26. Users can trigger the assistant with the "Hey Plex" wake word for complex research.
8)
@virtuals_io launched Fabric Protocol ($ROBO) as its first "Titan" project, focusing on the robot economy. Fabric provides identity and payment infrastructure to enable robots to function as autonomous economic participants.
9)
@LitProtocol introduced v3 Chipotle, a rebuild of its programmable signing and compute network. The architecture decouples execution from key management to provide AI agents with native, HTTP-based cryptographic workflows.
10)
@SentientAGI launched Arena, an evaluation platform for AI agents, with partners including Founders Fund and Pantera Capital. The platform benchmarks agent performance on real-world enterprise reasoning tasks to advance state-of-the-art.
11)
@sreeramkannan published the EIGEN 2026 letter, prioritizing verifiable compute for the agentic era. EigenCloud aims to provide full-stack verifiability for autonomous agents through its Data, Compute, and AI primitives.
12)
@vana released DataConnect, an open-source desktop app and SDK for exporting personal data from platforms like ChatGPT and Spotify. The tool enables builders to integrate portable user context into AI applications.
13)
@ethos_network launched Human Verification on
@base to combat automated social media traffic. The protocol uses onchain social consensus and slashing mechanisms to verify reputable human users without invasive biometrics.
14)
@omarsar0 highlighted new research from Sakana AI introducing Doc-to-LoRA. The hypernetwork instantly internalizes long documents into compact LoRA adapters in a single forward pass, drastically reducing inference latency and memory costs.
15)
@SuccinctLabs released a study showing that seven leading AI image detectors failed to identify manipulated images after minor edits. The researchers argue for cryptographic provenance over detection to prove authenticity.
16)
@ValeoCash released Payment Router, an x402 primitive that enables AI agents to orchestrate multi-provider payments. The tool enforces budget caps and issues a single unified cryptographic receipt for complex agentic workflows.
17)
@handshake_58 shipped five integrations to enhance agent autonomy on Bittensor, including VPN purchases and web scraping payments. The updates allow agents to transact seamlessly for data and infrastructure within execution loops.
18)
@simononchain noted that
@openclaw, recently acquired by OpenAI, recommends
@AskVenice for privacy-first inference. Venice allows agents to access models via
$VVV staking, providing compute at zero marginal cost.