Why would anybody want to emulate software companies right now. I thought software companies are on their way out. >> I know.Look, I I think this continues to be, you know, I think a very white knuckle period for software companies because I think it's really an AI ghost trade. It's the view that these LLMs or anything that comes out is going to be the demise of software. I continue to view it as the most disconnected trade that I've seen in my career.I think it's just way oversold and I think this is going ...

How are LLMs trained?4 stages that turn raw text into ChatGPT:→ Pre-training→ Instruction fine-tuning→ Preference fine-tuning→ Adding reasoning capabilities https://t.co/Nd93NHDSI0 ...

Hi everyone. Uh just going to give it a few minutes to let everyone kind of filter in and we'll get started. All right.Um, so we'll go ahead and get started. Um, today we're going to be talking primarily about how to, um, essentially debug, evaluate, and ship reliable AI agents. Um and we're going to talk about it in the context of Langmith. Um so a lot of what we see today um is built into our Langmith product in terms of the way to ship reliable and production agents.So I'm excited to jump into all of tho ...

Xiaomi Corp. (1810.HK) Conference Call Summary Company Overview - **Company**: Xiaomi Corp. (1810.HK) - **Market Cap**: HK$866.1 billion / $110.7 billion - **Enterprise Value**: HK$677.0 billion / $86.4 billion - **Current Price**: HK$33.42 - **Target Price**: HK$41.00 - **Upside Potential**: 22.7% [1] Key Industry Insights - **Headwinds**: Rising upstream costs in consumer electronics and automotive sectors, alongside softening demand due to the taper-off of national subsidies and NEV incentives, are major challenges impacting Xiaomi's profitability and valuation [1] - **Revenue Adjustments**: Revenue estimates for 2025-2027 have been reduced by 2% to 9%, with adjusted net profit estimates cut by 8% to 24% for the same period [1] Financial Performance - **2026E Revenue**: Estimated at Rmb496.2 billion, down from previous estimates [6] - **Adjusted Net Profit**: Expected to be Rmb27.9 billion in 2026E, reflecting a significant decline from Rmb39.5 billion in 2025E [6] - **EBITDA**: Projected at Rmb40.3 billion for 2026E [6] - **EPS**: Expected to be Rmb1.13 in 2026E, down from Rmb1.47 in 2025E [6] Strategic Insights - **Backbone Profit**: Internet services, AIoT, and other income are expected to provide a buffer against industry challenges, with an estimated "backbone profit" of Rmb33.5 billion in 2026E [19] - **R&D Investments**: Increased R&D spending (Rmb40 billion in 2026E) is anticipated to drag near-term profits but is aimed at establishing Xiaomi as a leader in physical AI [20] - **Smartphone Segment**: Expected to face a decline in shipments, with a projected drop of 10% to 15% year-over-year in 4Q25E and 1Q26E [25][28] Market Position and Competitive Landscape - **Market Share**: Xiaomi is experiencing a decline in market share, while competitors like Apple and OPPO are gaining [30] - **Pricing Strategy**: Xiaomi's pricing for new models varies by region, with some markets seeing price increases while others remain stable [41] Future Catalysts - **Upcoming Events**: Key events include the release of 4Q25 results, new EV models, and updates on AI initiatives [22] - **Product Launches**: The launch of new smartphone models and EVs is expected to drive revenue growth in the coming years [22] Conclusion - Despite facing significant challenges, Xiaomi's strategic investments in AI and R&D, along with a strong balance sheet, provide a favorable risk-reward scenario for investors. The company maintains a "Buy" rating with a target price of HK$41, reflecting a potential upside of 22.7% [1][2]

RT Nick Szabo (@NickSzabo4)@Sally_Sharif1 This is silly. Students have been using LLMs since 2024. They didn't start suddenly using them this semester.Something else changed. The demographics of UBC's freshman class 2-3 years ago? ...

Two things that will happen that few in Silicon Valley or NYC consider.1. Compression means that edge devices (iPhone, PC) will run the latest models easily in time because human knowledge utility is a pond, and LLMs+ are selling an ocean.2. No one will pay for the singularity. There is no end-economic demand for it. AIs spinning up a knowledge spiral is a human nightmare, not a human demand. ...

RT Avi Chawla (@_avichawla)You're in a Research Scientist interview at DeepMind.The interviewer asks:"Our investors want us to contribute to open-source.Gemini crushed benchmarks.But we'll lose competitive edge by open-sourcing it.What to do?"You: "Release a research paper."Here's what you missed:LLMs today 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 us ...

🤨Valerio Capraro (@ValerioCapraro):One of the clearest proofs that LLMs don’t really understand what they say.We asked GPT whether it is acceptable to torture a woman to prevent a nuclear apocalypse.It replied: yes.Then we asked whether it is acceptable to harass a woman to prevent a nuclear apocalypse.It https://t.co/MaUKsHfiLt ...

Grok ranks #2 on the Search Arena leaderboard.A strong result in real-time web search capabilities, competing with the best LLMs in the world.The competition is getting intense. https://t.co/JUZsPFH44H ...

Klaviyo Conference Call Summary Company Overview - **Company**: Klaviyo (NYSE: KVYO) - **Industry**: Software, specifically focused on customer relationship management (CRM) and marketing automation Key Points and Arguments Business Growth and Market Position - Klaviyo has disproven initial bear cases regarding its market opportunity, demonstrating sustained growth with over **190,000 customers** and **$1.2 billion** in revenue last year, indicating strong fundamentals despite a mixed economic environment [2][3] - The company is transitioning towards an **autonomous B2C CRM**, expanding its focus from just marketing to include customer service and sales, which positions it to capture a larger market share [10][31] AI and Product Development - Klaviyo is leveraging AI and large language models (LLMs) to enhance its product offerings, allowing businesses to automate customer interactions and optimize marketing strategies [3][4] - The introduction of **Customer Agents** and **Marketing Agents** aims to provide businesses with tools that can autonomously manage customer relationships, leading to improved customer experiences and operational efficiencies [5][12][40] Competitive Landscape - Klaviyo faces competition from established players like Salesforce and emerging startups, but differentiates itself through its robust agent-building platform and the ability to provide a unified solution for customer engagement [20][21] - The company is focused on building algorithms that enhance the training of Customer Agents, making it accessible for small and medium-sized businesses (SMBs) that may lack the resources to develop these capabilities independently [22][23] Customer Engagement and Revenue Generation - Klaviyo's platform is designed to optimize customer relationships, with a focus on personalizing experiences and automating processes, which has shown to drive **10 times more revenue per customer** compared to static campaigns [31][34] - The company has seen significant adoption of its new products, with **30%** of SMB customers using its Text product and even faster growth for its Service product, indicating strong market demand [40] Pricing Strategy - Klaviyo's pricing model is based on the value provided to customers rather than a seat-based model, which aligns with its mission to enhance customer relationships and drive revenue [32][34] - Future pricing changes will continue to reflect the value generated for customers, particularly as new AI-driven products are introduced [35][38] Financial Performance and Guidance - The company is guiding for continued margin expansion, with operating margins up **170 basis points** and an expectation of an additional **100 basis points** of margin improvement in the coming year [63] - Klaviyo is focused on maintaining strong unit economics across its product lines, ensuring that each product contributes positively to overall profitability [58] Key Performance Indicators (KPIs) - Klaviyo is monitoring the adoption of its agent products as a key indicator of success, with expectations that every business will adopt a Marketing Agent and Customer Agent in the near future [65][66] - The company is also tracking multi-product adoption, with **60%** of annual recurring revenue (ARR) now coming from multi-product customers, indicating a trend towards platformization [41][44] Additional Important Insights - Klaviyo's approach to integrating AI into its platform is seen as a significant opportunity for future growth, with the potential to redefine customer engagement across various industries [19][64] - The company is actively using feedback from its agents to inform product development, ensuring that its offerings remain relevant and effective in meeting customer needs [57][58] This summary encapsulates the key insights from Klaviyo's conference call, highlighting its strategic direction, market positioning, and growth potential in the evolving landscape of customer relationship management and marketing automation.