Cody Pawlowski

Great article on the approach to agentic architecture. A lot of similarity on how we've been developing some of our experiments recently. A hybrid of code + LLM logic and orchestration. https://lnkd.in/g84Eg5Pw

17

The Stacks. TLDR. (<2 min read) Sandboxes and bad questions We opened this month's Stacks by sharing our new AI integration, why Sandboxing is the right choice for our database and why this is the exact feature that you didn't ask for, but needed. Our second AI editorial shares the importance of asking good questions when doing research and how better search strategies won't completely save us. Fibro: Make a Decision Already Our fibromyalgia content gives you the confidence push you need to stop referring out these patients and making the freaking diagnosis yourself. Standards of care guidelines call for primary practitioners to just call it already. If you use the current guidelines, you're not likely to flag the wrong person. We promise. ED in men (and not the one you're thinking) Eating disorders are maybe more common in your patients with ovaries, but the consequences of missing it in the other 50% of your patients is just as serious. We share how to flag your patients with eating disorders regardless of gender, and why you need to be more careful looking at every patient that sits across from you. Thumbs down for B3 on cholesterol In our sneak peek content (typically for members' eyes only), we share the importance of clinically relevant outcomes for cholesterol care and how niacin doesn't fit the bill for cholesterol lowering. Hashimoto's and Ovarian Reserve A few important papers were published in the last 18 months creating even more of a concrete connection between HT and POI. If you're working with people TCC who have a positive TPO, you may want to read this ASAP. Want to read the whole thing? Click here: https://joom.ag/hStd

2

As LLM's get increasingly more sophisticated, with OpenAI's o1 model now including reflection before answering, I am increasingly investigating the emergence of autonomous AI Agents. My path here was somewhat circuitous in that I was going through Y Combinator's Summer 24 cohort which had a plethora of AI use cases that were very industry specific. I found that curious in that earlier SaaS offerings should have already automated and/or digitized most of these process, and I couldn't figure out why AI startups would focus on basic workflows such as inventory management and book keeping. As I dove into the matter, with the help of my now indispensable Perplexity, I landed on a very insightful post by Bessemer Venture Partners (https://lnkd.in/gHtQQRha). The thesis is that Vertical AI companies are going to be magnitudes bigger than general solutions, especially as they focus on industries that were earlier inaccessible to SaaS given very high language based workflows. Within that thesis, there was also a mention of AI Agents and how these would come to dominate workflow automation going forward. This neatly tied in with a great conversation between Peter H. Diamandis and Emad Mostaque (https://lnkd.in/g-ccxbma) that talked about 10 billion agents!! I have spent the last few weeks diving into AI Agents with the help of content from interesting voices such as Adam Silverman (Hiring) and David Ondrej. Over the next few months, I will try to create and deploy AI Agents on some basic use cases to get hands on experience on these systems. In the meantime, I will close this post with a link to a video from Andrew Ng, a pioneer in AI systems that provides a short overview of AI Agents (https://lnkd.in/g8v4pUQu). I hope you enjoy the content and wish everyone a productive week ahead.

8

Alexander Petros (https://lnkd.in/gXZBxzTj) delves into htmx (https://htmx.org/), lightweight protocols, and open-source software's impact on the web's future. Listen at 🎙️ https://lnkd.in/gAf8Chxp

3

AI isn't just about replacing humans—it's about empowering them. 💪 AssuranceIQ, an innovative insurance startup, is pioneering a groundbreaking approach to leveraging generative AI: 1. They're using Large Language Models (LLMs) to score customer conversations. 2. This AI-powered system structures unstructured data from interactions. 3. The goal? Enhance the performance of human sales and support teams. 4. It's not about AI vs. humans, but AI amplifying human capabilities. 5. This could revolutionize how we train and develop customer-facing teams. This use case showcases the true potential of AI in business: not as a replacement, but as a powerful tool to augment human skills and drive performance. What other innovative ways could we use AI to empower rather than replace human workers? #AIInnovation #CustomerExperience #FutureOfWork #BusinessAI #TechTrends Share your thoughts on AI augmentation in the comments below!

10

1 Comment

Generally surprised to not see more hardware vendors in healthcare and life sciences partnering with AI-first companies. When the zone of genius of a company is hardware, it's difficult to imagine that same team achieving parity on an AI product – Apple may be the greatest example of success in this realm, and still has secured many AI capabilities via targeted M&A (Siri, Turi, Xnor.ai, Voysis, Inductiv). Given the importance of leveraged distribution models in the ossified industries of healthcare and life sciences, it seems there is untapped opportunity for more hardware manufacturers to partner and co-develop AI products with startups. Control over the way data is captured and processed is critical for building high quality training data sets, which creates an interesting feedback loop between hardware companies and potential AI-first startup partners. Interested to explore this further with folks in healthcare and life sciences hardware!

150

10 Comments

The July 25th Open-Source LLaMA 3.1 release deserves attention, especially to the investors of close source LLM models. 3 model sizes of 8B, 70B and 405B parameters enables edge to datacenter usage. Expanded context windows to 128k tokens supports much large prompt and multilingual support. The 405B parameter model competes well with GPT-4 and Claude 3.5 and is integrated with various cloud platforms. The LLaMA model can be fine-tuned and adapted for specific tasks without full retraining with techniques like LoRA as full retraining requires 2048 A100 GPUs for 21 days (estimate). The hardware costs alone for the training could reach tens of millions of dollars (painful to VCs) so efficient alternatives to full retraining is key to AI startups leveraging LLaMA. The approaches are growing, but include LoRA, 8-bit precision to reduce memory footprint and fine-tuning on specific datasets or tasks. Why did Meta open-source LLaMA? Meta aims to disrupt the competitive edge of companies with proprietary models such as Google and OpenAI. The traditional open-sourcing benefits of building a large developer base, fostering innovation and adoption. It also helps Meta attract and retain top AI researchers who prefer working with open and accessible technologies. Likely, Meta will offer managed services with specialized hardware in the future.

16

2 Comments

July 29 2024 – 5 Notable Articles To Read Article: NVIDIA Partner SMC Raising $950M to Tap AI Server Boom Takeaway: SMC - Sustainable Metal Cloud provides data center capacity to clients, offering them access to Nvidia AI chips and advanced cooling technology. Source: Bloomberg Link: https://lnkd.in/dFXr43uT Article: OpenAI Is Launching Search Engine, Taking Direct Aim at Google Takeaway: Developed with input from publishers, SearchGPT will provide real-time summaries of website information. Source: The Wall Street Journal Link: https://lnkd.in/dSDj5fpA Article: Royal Caribbean Group Raises Its Outlook. Cruise Demand Is Strong. Takeaway: “Exceptional demand for our vacation experiences has accelerated our performance by generating significant yield growth over the past several years,” Chief Executive Jason Liberty said in the earnings release. Source: Barron's Link: https://lnkd.in/dpznkWbS Article: Alibaba Group-backed Baichuan raises $691M as China seeks to catch up on generative AI Takeaway: The business is valued at 20B yuan with the funding round included existing backer Alibaba, along with internet giant Tencent and smartphone maker Xiaomi Technology. Source: PitchBook Link: https://lnkd.in/dmWTKFan Article: J.P. Morgan pitches in-house chatbot as AI-based research analyst Takeaway: Approximately 50,000 employees are using a large language model developed by the bank to enhance productivity. Source: Financial Times Link: https://lnkd.in/dx6Z2sU4 #news #ai #research #business

5

Labor budgets are 35x software budgets, and AI agents eat into both. A few takes on short-term impact: 𝗛𝘂𝗺𝗮𝗻 + 𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗳𝗶𝗿𝘀𝘁: In the short term, a "human + machine" angle still wins. Industries with built-in failover mechanisms (like customer support) are ideal for initial AI deployment, allowing AI to handle basic tasks now and more complex ones over time. 𝗥𝗲𝗶𝗻𝘃𝗲𝗻𝘁𝗶𝗻𝗴 𝗣𝗿𝗶𝗰𝗶𝗻𝗴: AI agents challenge traditional SaaS models by moving beyond point-and-click interfaces, siloed data, and seat-based pricing. 𝗘𝘅𝗽𝗮𝗻𝗱𝗶𝗻𝗴 𝗔𝗰𝗿𝗼𝘀𝘀 𝗠𝗮𝗿𝗸𝗲𝘁𝘀: AI agents are transforming industries by deploying across horizontal applications (cross-industry tools like legal AI for enterprises), vertical applications (industry-specific solutions like AI for law firms), and consumer services (full-stack services like legal assistance apps). 𝗤𝘂𝗲𝘀𝘁𝗶𝗼𝗻𝘀 𝗼𝗻 𝗺𝗼𝗮𝘁: Rapid advancements in foundational AI models may commoditise a lot of early / short-term winners' use-cases. Raises questions on what sectors and verticals are more suitable to build product moats around. AI’s accuracy in regulated industries is improving (these plays could have a significant moat). Credit + a couple great reads / listens: - The Agent Economy by Felicis - The All-In Podcast (21/9/24) #ai #tech #llms #founders #startups

145

24 Comments

Time Inc. recognized Cohere's Command R+ model among the best inventions in 2024 for business-centric AI. Congratulations team! "Cohere's Command R+, an AI large language model released in April, excels at tasks businesses need: summarizing documents, answering questions based on that information, and responding in multiple languages. It also has the ability to interact with apps; for instance, Command R+ can use a calculator app to solve a math problem or an email client to share a document." #ai Thomvest

18

We're trying a new thing: Company spotlights to help people better understand what various health IT vendors actually do 🔦 We’ll be cutting through the marketing speak to give short, to-the-point descriptions of a company’s product and/or service + some context on how they see their differentiators. Tag any companies in the comments you'd like to see us cover (and follow Elion if you want to stay in the loop!)

35

13 Comments

I'm not sure we could navigate modern life without the tech that allows asynchronous communication. But it comes at a price. For example, with multiple tools like digital whiteboards, cloud-based storage systems, video recording, AI note-taking, email and chat, teams can work together even when they are distanced both by geographic location and by time. The tradeoff is that there must be broad agreement on and fastidious adherence to guidelines about how and when each team member will use each tool. If they can't be together physically, they'll need to be together systemically. Putting these systems in place and creating a culture of using them properly is an ongoing battle. How is your team managing it? #async #smb #remoteworklife #worklife #management #leadership #productivity https://lnkd.in/gTWhiaUN

8