RT Avi Chawla (@_avichawla)Deploy and run LLMs directly on your phone!Unsloth now lets you fine-tune LLMs and deploy them 100% locally on iOS/Android devices.The video shows this in action, where I ran Qwen3 on an iPhone 17 Pro at ~25 tokens/s.I have shared a guide in the replies. https://t.co/p4NqLj0jRE ...

Deploy and run LLMs directly on your phone!Unsloth now lets you fine-tune LLMs and deploy them 100% locally on iOS/Android devices.The video shows this in action, where I ran Qwen3 on an iPhone 17 Pro at ~25 tokens/s.I have shared a guide in the replies. https://t.co/p4NqLj0jRE ...

RT ashen (@ashen_one)I vibe-coded a website using @Replit that allows you to use x402 payments on @Solana to access LLMs without an API keyPay per request with USDC, no API keys needed, only on Solana.Test it out here: https://t.co/k0BClamlR4(full video on how to do this soon) https://t.co/IZnOEXwr8M ...

AWS doubles down on custom LLMs with features meant to simplify model creation https://t.co/gpzghqnW4I ...

AI Development & Strategy - Microsoft's AI strategy shifts towards AI self-sufficiency [1] - Microsoft is pushing towards "humanist superintelligence" [1] - The company is forming a new superintelligence team [1] - The advancements in power, data, and memory are crucial for AI progress [1] Technological Aspects - Memory plays an important role in AI [1] - LLMs (Large Language Models) are significant in the path towards superintelligence [1] - Power constraints are a key consideration in AI development [1] - Data is crucial for AI development [1] Economic & Future Trends - The economics of frontier models are facing price pressure and commoditization [1] - The vision includes cheap and abundant intelligence [1]

Reasoning Performance of LLMs - LLMs' reasoning performance significantly reduces when reasoning is obfuscated using simple ciphers [1] - LLMs struggle to reason accurately in ciphered text, except for simple ciphers [1] Ciphered Language Processing - LLMs can easily decode ciphered text to English [1] - Research involved testing 10 LLMs across 28 ciphers [1]

Broadcom & OpenAI Deal - Broadcom's stock surged nearly 10% following the announcement of a major deal with OpenAI, marking the latest in a series of significant data center agreements [1] - Broadcom is investing in and enabling select customers to run their Large Language Models (LLMs) due to a shortage of compute capacity [3] - OpenAI's valuation is over $500 billion, with strong revenue growth, making them a worthwhile partner for Broadcom [10] Power & Infrastructure - The power needed for compute is substantial, potentially requiring the equivalent of 18 Hoover Dams [4] - Usable power is key, with a shift towards distributed power sources of 50-200 megawatts, rather than relying on single large sources like 1-2 gigawatt facilities [6] - Companies like Oracle, Google, Meta, and Microsoft have access to available power, but making these sites usable is the challenge [5][6] Broadcom's Strategy & Market Position - Broadcom focuses on a select group of players creating "super intelligence" or generative AI, emphasizing customized silicon at scale [6] - Broadcom has been working on custom AI accelerators for eight years, starting with Google, and continues to learn and engage deeply [6] - Broadcom defines customers as those who provide production purchase orders at scale, indicating a hard-nosed business approach [12] - The demand for compute capacity is doubling every year, creating ample opportunity for multiple players in the market [17] - VMware, acquired two years ago, is performing extremely well, generating significant free cash flow while growing [18] Generative AI Outlook - Generative AI is viewed as a critical utility for society and a secular trend, not a cyclical one [19][20] - Knowledge-based technology industries account for 30% of global GDP (approximately $110 trillion), and generative AI could increase this to 40%, representing a $10 trillion annual growth [21]

Industry Overview - The AI industry is experiencing rapid developments, with significant investments and deals being announced frequently, indicating a rush in the market [1][3] - There is a total of $15 trillion in foreign direct investment into the US to support AI initiatives, alongside a market cap of $25 trillion in the top 10 tech companies [3] Company Dynamics - Companies are engaging in a circular economy, where chip companies are acquiring AI firms and vice versa, raising concerns about the sustainability and outcomes of these investments [5][6] - OpenAI is central to the current AI landscape, with significant stock holdings in AMD, indicating a strategic partnership in AI infrastructure [4] Market Trends - The market is currently in a transitional phase between the internet age and the AI age, with ongoing developments in large language models (LLMs) and chip technology [7][9] - Major tech companies like Salesforce and Oracle are positioning themselves to adapt to the evolving AI landscape, with Salesforce focusing on reinventing itself and Oracle showcasing advancements in AI at their upcoming event [12][14] Financial Insights - There is a substantial amount of capital flowing into the AI sector, with trillions needed for ongoing projects, sourced from private credit and capital expenditures [6][10] - The individual performance of tech companies has been remarkable, with significant financial gains being realized in the market [10][11]

RT Avi Chawla (@_avichawla)You're in a Research Scientist interview at OpenAI.The interviewer asks:"Our investors want us to contribute to open-source.o3 crushed benchmarks.But we can lose a competitive edge by open-sourcing it.What do we do?"You: "Release the research paper."Interview over.You forgot that LLMs don't just learn from raw text; they also learn from each other.For example:- Llama 4 Scout & Maverick were trained using Llama 4 Behemoth.- Gemma 2 and 3 were trained using Gemini.Distillation helps ...

Core Insights - The demand for compute capacity is accelerating, particularly for training large language models (LLMs) [1] - Broadcom's acquisition of VMware has positioned it as a significant player in enterprise software, but its primary focus remains on infrastructure, with semiconductors contributing over half of its $16 billion revenue in Q3 2025 [2] - The shift towards AI compute has significantly benefited Nvidia, with revenues reaching $46.7 billion, marking a 56% increase year-over-year [4] Investment Trends - Major cloud providers are significantly increasing their capital expenditures, with Oracle raising its annual plans to over $35 billion, Google Cloud adding $10 billion to its $75 billion budget, and Microsoft and AWS planning $100 billion and $80 billion respectively [5] - Investments in data center semiconductors, servers, and storage components grew 44% year-over-year to nearly $80 billion in Q2 2025, driven by hyperscaler infrastructure buildouts [6] Vendor Strategies - Companies are focusing on providing infrastructure to support high-capacity workloads as generative AI becomes integrated into business functions [7] - Nvidia is expanding its focus to enterprise customers, providing hardware for generative and agentic AI workloads in hybrid environments [11] - Google Cloud has seen a 27-fold increase in usage of its BigQuery data cloud as enterprises prepare their data for AI [13] Market Dynamics - The server market has experienced a resurgence, with year-over-year sales surging 134% [20] - The AI PC market is currently sluggish, but HP has reported an uptick in shipments, indicating early signs of growth [21] - Intel is facing challenges in deploying AI infrastructure, having recently undergone leadership changes and layoffs due to stalled revenues [18]