BREAKING: We exposed how George Soros converts tax-deductible donations into political campaigns through a $34 million pipeline where Tides and Working Families Party coordinate to support candidates like Zohran Mamdani Our investigation uncovered a $2+ billion tax-exempt network that systematically converted charitable donations into political power. The numbers tell the story. FULL INVESTIGATION: [link]: ➡️https://t.co/93ZjYqcPOr THE BASIC SCHEME: George Soros donates to his Foundation to Promote Open Society (501c3) - tax deductible → $8.4M flows to Tides Foundation (501c3) → $29.3M flows to Tides Advocacy (501c4) for "political advocacy" → Coordinates with Working Families Organization supporting candidates like Mamdani THE SMOKING GUN - REVERSE FLOWS: After receiving $9.71M from the Soros pipeline, Working Families Organization sent $325,000 BACK to Tides entities. Independent charities don't do this. It's mathematical proof of coordination. SYSTEMATIC CONCEALMENT: Tides Advocacy (501c4) lists Janiece Evans-Page as Director, while she simultaneously serves as CEO of both Tides Foundation (501c3) and Tides Center (501c3), yet Tides Advocacy filed NO disclosure forms for 5+ years while receiving $29.3M. THE EIN SCANDALS: ➡️Working Families Organization reported $80,000 to "WFP National PAC" but used an EIN that actually belongs to "International Earthlight Alliance." ➡️Tides Center reported a $343,753 contribution to "Tides Advocacy" using an EIN that belongs to "Think Outside Da Block Inc. PROFESSIONAL AUDITOR CONFIRMATION: Deloitte auditors confirmed operational integration between Tides Foundation and Tides Center, requiring consolidated financial statements. Yet Tides Advocacy - sharing the same leadership - systematically concealed these relationships. 6/ THE SCALE: ➡️$34.4M in documented flows between supposedly independent entities ➡️$16.1M in government grants flowing to organizations with political arms ➡️$36.5M in admitted "political transfers" by Open Society Institute ➡️Over $2B in total network operations OUR EVIDENCE: Every dollar amount traced to original IRS Form 990 filings. We've filed a Fifth Supplemental IRS whistleblower complaint documenting potential $500M+ in tax recovery from this coordination system. THE MAMDANI CONNECTION: This network manufactured his "grassroots" campaign while maintaining the illusion of independent charitable work. The infrastructure that created his political rise operates continuously, creating candidates nationwide. WHY IT MATTERS: If charitable tax deductions can be systematically converted into political coordination through concealed control structures, it undermines the entire tax-exempt system. Donors get tax breaks, candidates get "grassroots" cover, taxpayers subsidize political operations. The mathematics don't lie. When organizations engage in $34M+ flows while concealing key relationships and sending money backward through the system, these patterns warrant investigation.

Grok 2 is now #1 trending on Hugging Face 💫 https://t.co/ae4Ixk383d

We're announcing the Anthropic National Security and Public Sector Advisory Council, a bipartisan group of defense, intelligence, and policy experts who will help us support the U.S. government and closely allied democracies in maintaining our AI leadership. https://t.co/Yl6dYvE2dS

Members have collectively led major intelligence agencies, directed nuclear security operations, and shaped national technology strategy at the highest levels of government. Read more: https://t.co/ciRMIIOWPS

you can now mess with font size and shut dithering off on nous chat as well :) https://t.co/EBC9RFt9n1

you can now mess with font size and shut dithering off on nous chat as well :) https://t.co/EBC9RFt9n1

hermes-4 looks cute https://t.co/nFxI0rIqQW

hermes-4 looks cute https://t.co/nFxI0rIqQW

@janekm @opdroid1234 @ethanmlam @NousResearch Umm we did publish it and all that info is in the paper? Benchmark Code: https://t.co/UAXolOpcfN Dataset for evaluating with: https://t.co/xLMhxlDd2c

I tried out the new Nous Hermes AI thingy this things on an whole another level the first time I got impressed by an Ai LLM was Claude 3 Sonnet and now this when It Roleplays it just comes to life https://t.co/8iDrW3VsRH

Me emerging from the data mines after spending the last 3 months battling all the synthetic data kobolds to get Hermes 4 together https://t.co/8Yy5y9ImVT

@HighyieldHarry Anthropic CEO talked about how he approaches LLM economics in an interview with John Collison https://t.co/2d6SF9mb7L

Never take option #3 on humanity's annihilation. If you neglect #4 you will end up believing insane things as your mind thrashes around looking for ways to be okay. If people neglect #1 we all die. Contemplate #2 monthly but not daily. https://t.co/l0nzRyCxpp

6 months ago https://t.co/g83VzyMGsg

The raw truth: across 16 leading LLMs, AI agents succeed at fewer than 1 in 4 tasks. These tests use simple API calls, revealing a costly, fragile system prone to failure outside demo environments on current MCP servers. https://t.co/cIKr7BWGzT

Build better RAG by letting a team of agents extract and connect your reference materials into a knowledge graph. Our new short course, “Agentic Knowledge Graph Construction,” taught by @Neo4j Innovation Lead @akollegger, shows you how. Knowledge graphs are an important way to store information accurately but they are a lot of work to build manually. In this course you’ll learn how to build a team of agents that turn data– in this case product reviews and invoices from suppliers–into structured graphs of entities and relationships for RAG. Learn how agents can automatically handle the time-consuming work of building graphs — extracting entities and relationships (e.g., Product "contains" Assembly, Part "supplied_by" Supplier, Customer review "mentions" Product), deduplicating them, fact-checking them, and committing them to a graph database — so your retrieval system can find right information to generate accurate output. For example, you can use agents to help trace customer complaints directly to specific suppliers, manufacturing processes, and product hierarchies, thus turning fragmented information into queryable business intelligence. Skills you’ll gain: - Build, store, and access knowledge graphs using the Neo4j graph database - Build multi-agent systems using Google’s Agent Development Kit (ADK) - Set up a loop of agentic workflows to propose and refine a graph schema through fact-checking - Connect agent-generated graphs of unstructured and structured data into a unified knowledge graph This course gets into the practicum of why knowledge graphs give more accurate information retrieval than vector search alone, especially for high-stakes applications where precision matters more than fuzzy similarity matching. Sign up here: https://t.co/2txZfYqGZ9

Is the AI boom finally starting to slow down? https://t.co/230p5q58Xs

Melania Trump launches AI contest for schoolchildren in grades K-12 https://t.co/MEGlqsLPnk

How AI Echo Chambers May Be Fueling Psychotic Episodes https://t.co/eszHzNRIyT @ConorFeehly @sciam

Crypto's U.S. Policy Aims May Pivot on Resistance from Democratic Senator Warner https://t.co/nyUN549C6Q @jesseahamilton @coindesk

Meta Superintelligence Labs is already losing key staff less than two months after launch https://t.co/zT6S9v5Fom @PranavDixit @CharlesRollet1 @jyoti_mann1 @businessinsider

Behind the controversial AI tech used to inspect rental vehicles for damages https://t.co/v2Y6YtA4lT @cbseveningnews @krisvancleave @CBSNews

Anthropic Settles High-Profile AI Copyright Lawsuit Brought by Book Authors https://t.co/Dm0lSwvqI1 @Knibbs @wired

JPMorgan, Already No. 1 in Card Spending, Wants Even More https://t.co/IvvFu268Rc @wsj

Google Advances Its Layer-1 Blockchain; Here’s What We Know So Far Rich Widmann, head of Web3 at Google, outlined on Tuesday how his firm's upcoming layer-1 blockchain for finance differs from Stripe's Temp and Circle's Arc. https://t.co/FmAMISnojc @coindesk

Epic's ‘large medical AI model’ aims to change how hospitals predict patient risk https://t.co/1OyeabjtEL @brittanytrang @statnews

Early this summer, OpenAI and Anthropic agreed to try some of our best existing tests for misalignment on each others’ models. After discussing our results privately, we’re now sharing them with the world. 🧵 https://t.co/OMAJe3ZfQ8

🇺🇸🇮🇳 Trump’s 50% tariffs on India take effect The U.S. raised tariff rates from 25% to 50%, citing India’s purchases of Russian oil. Five rounds of negotiations between Washington and New Delhi failed to reach an agreement. https://t.co/3Q5eECp4lx

Everyone loves to hate Taylor Lorenz but her ability to find something that makes me laugh for a solid 10 minutes remains unmatched. The rest of you are too serious and simply not funny… https://t.co/Nm9a6VO6lK

I tried to use several “build your mobile app with AI” tools I haven’t been able to get any of them to work for me, even though Lovable, Replit, etc are decent So @isaac_flath and I are gonna give this one last go 2pm PST today You can heckle us here https://t.co/ahVZXup6E7 https://t.co/C1RwenNEFd

~ Ruler for Remote Async Agents ~ Example of running Ruler in a github action, so that your remote async agents (copilot swe agent, openhands, claude action, etc...) always have the rules available when you ask them to hack on something. https://t.co/UOSe2RxXK1