$spy levels for reference: gapped down, and breaking below 764.01. we're now waiting on bessent's upsized treasury operations around ~11am today, which should influence today's price action. PPI and CPI thursday/friday, can cause a retest of lower trend line ~760.40 LVN, 757.67 gap fill; or a potential breakout above 771-772 it's worth derisking shorter dates and waiting for confirmation

How to avoid money murdering market conditions: 1. Market gapped significantly overnight with no clear level to trade from 2. Price already hit the major weekly or daily level before the open 3. No clean support or resistance holding on the chart 4. No clear direction on $SPX going into the session 5. Major economic events or Fed announcements same day 6. Choppy price action with no follow through in either direction If any of these are present, do nothing. Remember: Cash is a position.

TRM's forward deployed engineering (FDE) team ships mission-critical capabilities into classified networks and air-gapped enclaves — environments with no internet access, no external AI tools, and direct lines to real investigations. The team is now expanding into applied AI in these environments. Get an inside look at how the team operates and what it takes to thrive here 👉

$SHAZ PARTNERS WITH VAST DATA ON CONFIDENTIAL AI INFRASTRUCTURE Sharon AI will integrate VAST DataEnclave into its AI Factory platform across Australia and Asia-Pacific, allowing regulated customers to run advanced AI models onshore, including in fully air-gapped environments. The setup keeps both customer data and model weights encrypted during processing, including GPU memory and interconnect traffic, with Nvidia GPUs running in Confidential Compute mode. Customers and model providers keep control of their own encryption keys, while even Sharon AI itself cannot access the underlying data or model weights. The platform is aimed at banks, government agencies and other organizations that cannot send sensitive workloads to conventional cloud-hosted AI services.

palantir and nvidia have stopped using claude/GPT when dealing with proprietary information due to concerns their data is being used to train a model that replaces them. instead they're using claude + gpt to automate generic HR, finance, legal work then shifting to open source + air-gapped private servers for confidential work (codebases, drug designs, software). anthropic, openai's argument is they need the data to prevent misaligned use.