David Alami

Another confirmation is rising Cumulative Volume Delta while the price was flat. More and more agressive buyers were entering. Each bounce off the whale's buy wall recruited more retail buyers. Then pull the sell wall → buy pressure + dried-up supply → pump.

But what was preceeding it? A crash with panic selling on high volume. Then a bounce and the volume died to nothing while the price stabilized. Volume dying while price holds = someone was absorbing all sell pressure silently. #PIPPIN

Most pumps follow this exact order: 1. Smart money accumulation 2. Liquidity positioning 3. Small price increase 4. Retail FOMO 5. Distribution (dump) Lets try to analyze the most recent pump of #PIPPIN

First step I think is to watch for unusual volume while the price is still ranging flat. Someone is accumulating quietly. They WANT it to look boring while they load up. What do you guys think is the best first signal to watch for?

Most of time I spot a pump when its too late, becoming an exit liquidity. Being tired of it, want to develop framework to detect orchestrated pumps BEFORE the chart goes vertical. #Crypto #SIRENUSDT #RIVERUSDT #DeFi #DYOR

#OpenAI launched GPT-5.1 Instant and GPT-5.1 Thinking for #ChatGPT, promising a faster, more conversational, and customizable AI experience after mixed feedback on GPT-5, with both models now available to most users

#baidu released ERNIE-4.5-VL-28B-A3B-Thinking, an open-source multimodal AI model that claims to outperform #Google and #OpenAI latest on vision benchmarks, all while flexing its efficiency by using just a fraction of the computing power

AWS and OpenAI have entered a multi-year strategic partnership, with OpenAI pre-paying AWS for cloud capacity in one of the largest AI infrastructure deals ever, aiming to diversify beyond Microsoft Azure and secure predictable revenue for AWS.

Magentic Marketplace launched an open-source simulation platform to study how AI agents interact and trade in digital markets, enabling researchers to experiment with agentic market designs at scale.

LangSmith launched a no-code Agent Builder in private preview, letting enterprises create agentic apps with Gmail, Slack, and Salesforce integrations while leveraging LangChain/LangGraph features

Google’s NotebookLM has supercharged its chat with an 8× larger context window, 6× longer memory, and goal-oriented prompts, all powered by #Gemini , promising 50% better answers for research tasks.

@AnthropicAI launched #Claude AI for finance, embedding it into Microsoft Excel and integrating real-time market data, aiming to rival Microsoft Copilot and capture a bigger share of the financial services AI market.

@OpenAI and @PayPal are teaming up to let ChatGPT users shop and pay instantly via PayPal right inside the chat, using Google’s Agentic Commerce Protocol

#OpenAI unveiled its grand plan to build a fully automated AI researcher by 2028, with @sama promising a junior AI assistant by 2026 and teasing a trillion-dollar infrastructure investment to make sure the robots have enough power to overthink everything.

Chinese startup MiniMax has released MiniMax-M2, an open-source #LLM that now tops global benchmarks for agentic tool use, offering enterprise-friendly licensing and compatibility with #OpenAI and #Anthropic APIs, challenging the proprietary giants

Traders There is only one legit reason for speculating in stocks/futures/crypto in my opinion That is to build wealth and maybe develop trading into a full time gig If your goal is to just have fun - then good luck. I have no time for you If your goal is do repeated 2X, 5X, 10X returns, then I hope you are trading zero sum markets because you will be my competition IMO, developing a long-term trading approach capable of consistent 20-40% returns without crazy asset volatility is a fabulous goal to have If this is your goal, then you have five major phases to go through Phase 1 - figure out what asset class is for you Phase 2 - learning how trading in that asset class works Phase 3 - developing an approach that can become your starting point for real-time forward testing and refining Phase 4 - Adopting an approach only after proving it to be profitable over several hundred trades Phase 5 - overcoming doubt and learning trust Battling ongoing is the biggest challenge career traders have -- every pro trader that has been fully transparent with me has admitted the same Battling doubt ongoing is basically what speculation is all about. Having doubt and second guessing is not a negative -- they come with the territory

Fibro-GPT was presented at #ACG 2025 as an open medical #LLM for liver-disease risk prediction, showing strong accuracy (AUROC 0.91) and promising transparent, reproducible research tools for healthcare AI.

Anthropic is investing in one million Google TPUs to supercharge its next-gen #Claude models, aiming for bigger context windows and faster #AI, while boasting 7X annual growth in big-spending business clients and a “diversified chip strategy” to stay ahead and manage costs.