Stefan Ebner

One of the sure ways to validate your startup is with a working business model It proves that there is a valid connection between your clients and your business In the file below you will find 31 Startup Business Models you must know (with brand examples) 👇 Credit: Herwig Springer & Navdeep Yadav Alvin Foo Debonkar Roy Pranali Jouras Linas Beliūnas

4

🆕 The Ultimate Guide to Prompting Sander Schulhoff and the Learn Prompting team reviewed hundreds of AI prompting papers to come up with the ultimate prompting taxonomy and strategies. If you want to get up to speed on all prompting techniques, this is the episode for you 🖋 [00:00:00] Introductions [00:07:32] Navigating arXiv for paper evaluation [00:12:23] Taxonomy of prompting techniques [00:15:46] Zero-shot prompting and role prompting [00:21:35] Few-shot prompting design advice [00:28:55] Chain of thought and thought generation techniques [00:34:41] Decomposition techniques in prompting [00:37:40] Ensembling techniques in prompting [00:44:49] Automatic prompt engineering and DSPy [00:49:13] Prompt Injection vs Jailbreaking [00:57:08] Multimodal prompting (audio, video) [00:59:46] Structured output prompting [01:04:23] Upcoming Hack-a-Prompt 2.0 project

27

8 Kommentare

Harvard Business School - AI Companions Reduce Loneliness Source: https://lnkd.in/g-tFvTmM #ai #companions #reduce #loneliness Credits: Julian De Freitas Ahmet K. Uguralp Zeliha O. Uguralp Stefano Puntoni Highlights: Generative AI is forecasted to grow into an impressive $1.3 trillion market by 2032, suggesting a concomitant rise in AI companion platforms. This is already evident in platforms such as XiaoIce, Chai, and Replika, among others. A user can ask their AI companion questions, and it will respond in a natural, believable way. The AI companion can also initiate conversations itself, such as “How are you feeling” or “Are you mad at me?” Consumers may use these platforms for both friendly and romantic purposes. For example, around 50% of Replika users have a romantic relationship with the AI. Here we consider the value proposition that AI companions reduce loneliness, inspired by an interview we conducted with the CEO of Replika and her investors. All suggested that consumers are using the app because they are lonely, and that the apps do in fact reduce feelings of loneliness. We make several contributions: 1. Research in psychology indicates that loneliness is a powerful emotional state that urges consumers to seek social connections. In the absence of available human interaction, individuals may engage with alternative forms of companionship, such as AI companion apps. We contribute to an understanding of whether loneliness motivates consumers to use AI companions, by detecting instances of loneliness in conversations on a real AI companion app and reviews of several such apps. We use Large Language Models (LLMs) that we fine-tune to detect loneliness, enabling us to detect loneliness more accurately than traditional dictionary-based approaches. We find, consistent with recent findings on LLMs, that only a small dataset of exemplars is required to specialize the LLM for our purposes, likely due to the rich contextual knowledge that the baseline models leverage for the fine-tuning task. 2. We explore whether conversations with AI companions help to alleviate feelings of loneliness, contributing to work on the efficacy of technological solutions like social robots in helping consumers cope with loneliness. In doing so, we study consumer loneliness both before versus after interacting with AI companions through textual conversation, at both cross-sectional and longitudinal scales. Thus, we also measure how consumers behaviorally interact with commercially representative versions of the technology, unlike most studies in the consumer behavior literature. 3. We contribute to understanding what features of chatbots lead to alleviation of loneliness, by systematically manipulating the conversational performance of the chatbot, and the ability of the chatbot to make the consumer feel heard—a construct involving the perception that the communication is received with attention, empathy, and respect

59

6 Kommentare

🅷 𝐁𝐨𝐨𝐤 𝐑𝐞𝐯𝐢𝐞𝐰: 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈 𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧 𝐈𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐏𝐚𝐭𝐭𝐞𝐫𝐧𝐬 "Generative AI Application Integration Patterns" by Luiz Lopez and Juan Bustos is a timely and insightful guide for professionals looking to integrate #GenerativeAI effectively into their business solutions. The authors take a pragmatic approach to the often-hyped world of #AI, emphasising the importance of aligning AI capabilities with tangible business challenges and workflows. 🅷 𝐊𝐞𝐲 𝐈𝐧𝐬𝐢𝐠𝐡𝐭𝐬 One of the book's standout quotes encapsulates its practical philosophy: ❝𝘜𝘯𝘧𝘰𝘳𝘵𝘶𝘯𝘢𝘵𝘦𝘭𝘺, 𝘵𝘳𝘦𝘢𝘵𝘪𝘯𝘨 𝘈𝘐 𝘢𝘴 𝘢𝘯 𝘦𝘯𝘥 𝘴𝘰𝘭𝘶𝘵𝘪𝘰𝘯 𝘳𝘢𝘵𝘩𝘦𝘳 𝘵𝘩𝘢𝘯 𝘢𝘯 𝘪𝘯𝘨𝘳𝘦𝘥𝘪𝘦𝘯𝘵 𝘵𝘰 𝘦𝘯𝘩𝘢𝘯𝘤𝘦 𝘴𝘰𝘭𝘶𝘵𝘪𝘰𝘯𝘴 𝘰𝘧𝘵𝘦𝘯 𝘭𝘦𝘢𝘥𝘴 𝘵𝘰 𝘧𝘢𝘪𝘭𝘶𝘳𝘦. 𝘞𝘦𝘭𝘭-𝘪𝘯𝘵𝘦𝘯𝘵𝘪𝘰𝘯𝘦𝘥 𝘵𝘦𝘢𝘮𝘴 𝘥𝘦𝘮𝘰 𝘧𝘭𝘢𝘴𝘩𝘺 𝘱𝘳𝘰𝘵𝘰𝘵𝘺𝘱𝘦 𝘤𝘢𝘱𝘢𝘣𝘪𝘭𝘪𝘵𝘪𝘦𝘴 𝘵𝘩𝘢𝘵 𝘧𝘢𝘪𝘭 𝘵𝘰 𝘮𝘢𝘱 𝘪𝘯𝘵𝘰 𝘵𝘢𝘯𝘨𝘪𝘣𝘭𝘦 𝘣𝘶𝘴𝘪𝘯𝘦𝘴𝘴 𝘤𝘩𝘢𝘭𝘭𝘦𝘯𝘨𝘦𝘴 𝘰𝘳 𝘸𝘰𝘳𝘬𝘧𝘭𝘰𝘸𝘴.❞ This perspective sets the tone for a book prioritising business value over technological showmanship. The authors consistently reinforce this idea, advising readers to: ❝𝘞𝘩𝘦𝘯 𝘦𝘹𝘱𝘭𝘰𝘳𝘪𝘯𝘨 𝘰𝘱𝘱𝘰𝘳𝘵𝘶𝘯𝘪𝘵𝘪𝘦𝘴 𝘵𝘰 𝘢𝘱𝘱𝘭𝘺 𝘨𝘦𝘯𝘦𝘳𝘢𝘵𝘪𝘷𝘦 𝘈𝘐, 𝘤𝘰𝘯𝘵𝘪𝘯𝘶𝘰𝘶𝘴𝘭𝘺 𝘦𝘷𝘢𝘭𝘶𝘢𝘵𝘦 𝘢𝘯𝘺 𝘱𝘰𝘵𝘦𝘯𝘵𝘪𝘢𝘭 𝘣𝘶𝘴𝘪𝘯𝘦𝘴𝘴 𝘷𝘢𝘭𝘶𝘦 𝘧𝘪𝘳𝘴𝘵 𝘳𝘢𝘵𝘩𝘦𝘳 𝘵𝘩𝘢𝘯 𝘫𝘶𝘴𝘵 𝘵𝘩𝘦 𝘵𝘦𝘤𝘩𝘯𝘪𝘤𝘢𝘭 𝘢𝘳𝘵 𝘰𝘧 𝘵𝘩𝘦 𝘱𝘰𝘴𝘴𝘪𝘣𝘭𝘦.❞ 🅷 𝐂𝐨𝐧𝐭𝐞𝐧𝐭 𝐚𝐧𝐝 𝐒𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 The book offers a comprehensive look at #integration #patterns for real-time and batch scenarios, providing readers with a versatile toolkit for various application needs. Its strong focus on practical implementation stands out, featuring numerous code examples that give readers concrete guidance on applying the concepts discussed. While the examples primarily use #VertexAI, the principles and patterns discussed will likely apply across different AI platforms. While this focus on Vertex AI might be a limitation for some readers, it also allows for deeper, more specific insights into a significant AI platform. 🅷 𝐓𝐞𝐜𝐡𝐧𝐢𝐜𝐚𝐥 𝐃𝐞𝐩𝐭𝐡 𝐚𝐧𝐝 𝐀𝐜𝐜𝐞𝐬𝐬𝐢𝐛𝐢𝐥𝐢𝐭𝐲 The authors have struck a balance between technical depth and accessibility. They provide a good overview of various integration patterns with generative AI, accompanied by implementation examples. This approach makes the book valuable for those new to AI integration and more experienced practitioners looking to refine their strategies. 🅷 𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐥𝐞 𝐀𝐈 𝐂𝐨𝐧𝐬𝐢𝐝𝐞𝐫𝐚𝐭𝐢𝐨𝐧𝐬 The book's inclusion of Secure and Responsible AI Frameworks is noteworthy. In an era when ethical AI use is increasingly important, this section adds significant value, helping readers navigate the complex landscape of AI ethics and security. Thank you, Anamika Singh and Packt, for sending the early review copy!

150

27 Kommentare

AI investment analysts are hitting the market! 🔥 Watch Firecrawl's LLM Extract build a database on the latest VC funding by pulling information from news articles. It provides structured details on investors, seed rounds, and more. Open-source repo in the comments!

41

6 Kommentare

UPDATED: For anyone who wants to access the Google Doc: https://t2m.io/hcgMnsHL [DOC] High Value Assets Documentation 30+ pages of AI prompts to speed-up High Value Asset creation for your company! It's simple: Copy-paste these templates into your preferred AI (e.g. ChatGPT, Claude) and receive Asset instructions in seconds. Then follow the AI output instructions to create Viral Assets for your LinkedIn page, website, newsletter, etc. Engage with your community and promote your brand with these bespoke visual messages. For example: - Cheat Sheets - Market Maps - Benchmarks - Flow Charts - Scorecards - Mind Maps - ... and much more for any industry These could save you hundreds of hours this year: >> Growth teams want these prompts to create marketing collateral. >> Sales teams want them to create sales enablement literature. >> Executives want them to build momentum. This Google Doc is one of the most valuable resources that we've ever shared publicly at GrowthMasters from our library of Standard Operating Procedures that have helped us ghostwrite B2B content generating 350M+ views on LinkedIn for our clients. Would love to share it with you! Want access? DM me "ASSETS" and I'll share the full Google Doc ASAP. . . BONUS: [PLAYBOOK] 6-Month Content Calendar: https://t2m.io/Ymrndxbw Support our startup/VC content by leaving a like, comment, or share on this post. Appreciate your support 🙌🙌 To your growth! 🚀 Note: Some of the finished image examples featured on this document were created by GrowthMasters.io, but not all examples were. Some have been curated from around the web. In those cases, all credit goes to its respective owners. However, all prompts were handcrafted by Juan Felipe Campos from GrowthMasters. #venturecapital #vc #b2b #growth #gtm . . P.S. 🔔 Follow me for posts on venture capital, startups, and M&A. 👉 Private Message me the word "PLAYBOOK" and I'll share a Content Engine marketing strategy for your own B2B LinkedIn page. 350M+ views generated: 🦁 GrowthMasters - Content for VCs, Startups, and Their Ecosystem. Clients have worked with a16z, Madrona, EY, Accel, and many more.

120

28 Kommentare

How to model an Earnout (skim in 30 seconds) 👇 ~~~ 📍𝗧𝗼𝗱𝗮𝘆'𝘀 𝗔𝗻𝗻𝗼𝘂𝗻𝗰𝗲𝗺𝗲𝗻𝘁: building a model but not sure where to start? Grab a free checklist here 👉 https://lnkd.in/eVTU67fY ~~~ 𝗪𝗵𝗮𝘁'𝘀 𝗮𝗻 𝗘𝗮𝗿𝗻𝗼𝘂𝘁? Earnouts help bridge a valuation gap between buyer and seller. For example: Buyer: "We think the business is worth [$X million]." Seller: "Well, 𝘸𝘦 think the business is worth [$X million + $50]." Buyer: "We believe you but need to see it, so let's do an Earnout. When the business reaches $500 million of TTM Adjusted EBITDA, we'll pay a $50 million Earnout." 𝟭. 𝗥𝗲𝗰𝗼𝗿𝗱 𝗘𝗮𝗿𝗻𝗼𝘂𝘁 𝗟𝗶𝗮𝗯𝗶𝗹𝗶𝘁𝘆 The Earnout is recorded as a contingent liability on the Balance Sheet. 𝟮. 𝗘𝗮𝗿𝗻𝗼𝘂𝘁 𝗴𝗼𝗲𝘀 𝘁𝗼 𝗚𝗼𝗼𝗱𝘄𝗶𝗹𝗹 Since the Earnout is potential consideration to the seller above-and-beyond the net asset value of the acquired company, it is recorded as Goodwill. (This way, the creation of the Liability and increase in the Goodwill offset to $0 for cash purposes in the model) 𝟯. 𝗞𝗲𝗲𝗽 𝗧𝗿𝗮𝗰𝗸 𝗼𝗳 𝗧𝗮𝗿𝗴𝗲𝘁 Earnout targets can be anything: - Revenue-based - Gross Profit-based - EBITDA-based Really though, it's anything you can think of: It is a performance target that needs to be hit. I've used a simple one in this example that "TTM Adjusted EBITDA needs to reach $500 million." 𝟰. 𝗥𝗲𝗱𝘂𝗰𝗲 𝘁𝗵𝗲 𝗟𝗶𝗮𝗯𝗶𝗹𝗶𝘁𝘆 Once the target is hit, the Earnout liability gets adjusted to zero and the cash leaves the business. If the company doesn't have enough cash to pay the Earnout, it would likely borrow from its debt facilities (like a Line of Credit). 𝗔𝗱𝘃𝗮𝗻𝗰𝗲𝗱 𝗦𝘁𝘂𝗳𝗳 (𝗻𝗼𝘁 𝗶𝗻 𝗽𝗶𝗰𝘁𝘂𝗿𝗲) Along the way, you can adjust the value of the Earnout liability based on the likelihood of it being paid out. Let's say you think there's only a 75% chance it will be paid (down from 100%), that 25% difference would reduce the liability and be recorded as a gain on the Income Statement, a non-cash expense. My personal 𝘱𝘳𝘦𝘧𝘦𝘳𝘦𝘯𝘤𝘦 is to leave this part out of the model and always keep the liability at 100% likelihood since (a) you can't predict the future and (b) I want my model to be conservative as possible. ⚠️𝗬𝗲𝗮𝗵 𝗯𝘂𝘁... 𝗱𝗼 𝘁𝗵𝗲𝘆 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝘄𝗼𝗿𝗸? My two cents: I think earnouts are genuine in their intent, but often push the management team too far and/or are unreasonable. I've probably seen ~10% go great, ~30% of them "go okay" and the other 60% not so much. So while it's not a "hard no," I'd be wary of any structure that is too "earnout-heavy" and make sure you're getting meaningful proceeds upfront that match the value you've already created. Either way, as a modeler, it's a concept you should understand. ~~~ 👋 Hey, I'm Chris Reilly, and I post about Financial Modeling in the middle-market. 𝘱.𝘴. don't forget to grab my free model checklist here 👉 https://lnkd.in/eVTU67fY

103

5 Kommentare

🚀 GPT-o1 (GPT-5 if you prefer 😉) , this isn't just another release—it's OpenAI's lifeline Key points for GPT-o1: - Prioritizes reasoning over speed (smarter, but slower) - 2 versions Mini and Full - No API access yet We are stress-testing it, we'll integrate where it makes sense, and give you the unvarnished truth. Why GPT-o1is OpenAI's all-or-nothing play: 👉 GPT-4 Lost its "one of a kind model" status: Anthropic Claude 3.5 and Meta Llama 3.1 are fully viable alternatives (Google Gemini is not far behind) . OpenAI's no longer the only AI powerhouse in town. 👉Brain drain : Summer saw a mass exodus of top talent. Not a great look for the future, not a great signal for customers and investors 👉Burning cash: They're hunting for $6.5 billion in funding. Why? Because cutting-edge AI devours money like no other tech before. With investor patience wearing thin and rivals closing in, they need a knockout, not just a jab. Prediction: Speed improvements are coming. They need to make this happen fast as this "smarter" model is too slow for conversational interfaces. What's your take on OpenAI's high-stakes gambit? Sound off below. 👇 #GPT5 #AITech #OpenAI #GPTo1

11

Venture Capital Methods of Valuation: Understanding valuation is crucial for startups navigating the venture capital landscape. This document provides a detailed overview of the different methodologies used to evaluate early-stage companies, helping founders and investors make informed decisions. Key Valuation Methods Highlighted: 1) Pre-Money and Post-Money Valuation: Learn how investments impact your company's valuation and how to calculate ownership dilution. 2) Scorecard Method: This method compares your startup to others in the sector, adjusting the valuation based on strengths and weaknesses in key areas such as management and market size. 3) Berkus Method: Focuses on qualitative factors, valuing the startup's potential based on the presence of a sound idea, prototype, quality management, strategic relationships, and product rollout. 4) Venture Capital Method: Calculates future returns based on projected revenues and exit possibilities, considering multiple rounds of funding and the effect of dilution on investor returns. 5) Sensitivity Analysis: Assesses how changes in key assumptions like terminal value, exit timing, and investment amount affect the overall valuation. PS. check out 🔔 for a winning pitch deck the template created by Silicon Valley legend, Peter Thiel https://lnkd.in/ejp-Bhnu

19

OpenAI just dropped a bombshell with GPT-4o I know everybody is talking about it for 24 hours. And come on it’s worth a talk because it's about to change the AI game forever (for all good reasons). Here's everything you need to know: → GPT-4o is an upgraded version of GPT-4 → It excels in text, vision, and audio processing → It's faster, cheaper, and more capable than ever → Text and image capabilities are rolling out today in ChatGPT → ChatGPT users get GPT-4o for free, with paid users getting 5x the capacity But that's not all. GPT-4o is natively multimodal. Now, what does it mean? This means it can generate content and understand commands in: ↳ Voice ↳ Text ↳ Images Can you even Imagine the possibilities this opens up? ChatGPT's voice mode is also getting a major upgrade. It will be like having a personal AI assistant, just like in the movie "Her". → Real-time responses → Going beyond single prompts → Observing the world around you OpenAI's vision has evolved over time, for sure.. And, instead of creating benefits directly, they're focusing on: → Creating powerful AI models → Making them accessible to developers through APIs → Enabling others to build amazing things with their tech This shift in strategy is a game-changer. It democratizes access to cutting-edge AI. In a way, it empowers creators and innovators worldwide. And it sets the stage for a new era of AI-driven innovation. The timing of this launch is no coincidence either. With Google I/O just around the corner, OpenAI is making a bold statement. They are ready to take on the tech giants head-on. The AI race is heating up, and OpenAI is leading the charge. Are you ready for the future of AI? P.S. What excites you the most about GPT-4o? #openai #gpt40 #techupdates #artificialintelligence

66

24 Kommentare

We are in the biggest bubble ever, one even bigger than the one that motivated the title of my book. The title was originally motivated by the startup bubble of 2021 that could be seen in record IPOs, venture capital funding, and share prices for publicly traded #startups. But the generative AI bubble has far eclipsed the startup bubble and may not even have reached its peak despite recent declines. Behind the bubbles: AI and other high-#tech startups have mistakenly assumed the new #technology would experience rapid improvements and quickly diffuse and thus they didn’t need to think carefully about first applications and customers. VR headsets aren’t getting smaller nor is field of view for AR goggles getting bigger while the problem of nauseousness still remains. Delivery drones cannot deliver packages to balconies of high-rise building, which dominate big cities, nor are they getting quieter or better at navigating power and telephone lines, a characteristic of suburbs. eVTOLs aren’t even being implemented. The biggest failure, however, lies in AI. After years of trying to replace radiologists, drivers, police officers, and home flippers, they are now focused on augmentation, a better goal, but also a difficult one for most applications. Augmented radiology is proceeding slowly, and transcription has become the most emphasized application in hospitals, a big step down from the claims a few years ago that AI would eliminate doctors and nurses. Self-driving vehicles still hallucinate providing cities with midnight honking and vehicles unresponsive to police officers, and yet supporters still claim they will eventually be safer than human driven vehicles. In the end, the real argument is about how fast improvements in AI and other new technologies will occur. We need to focus on the relevant variables when drawing conclusions about the improvements. Yes, improvements in image resolution are important when dealing with generative #AI for advertising videos, but for coding, reports, emails, and call centers, the frequency of #hallucinations is a more important variable and no matter how many new products are released, if that frequency doesn’t drop, generative AI will not be a major success. A big reason that many AI proponents are confused by the last paragraph is that they have fallen victim to the Wow! factor, one form of #hype. They believe in optimistic goals without asking how they will be achieved. The book is filled with these stories. One reason they fall for the Wow! factor is because the VCs and startups sold them the myths of genius entrepreneurs, frequent disruptions, and rapid improvements, which are debunked through an old technique called facts! Publication date is October 22. See the testimonials in the comment section. #innovation #science https://lnkd.in/gkgdwSuy

268

49 Kommentare

Goldman Sachs is now telling a pessimistic story about AI that is consistent with what David Cahn of Sequoia Capital (last night’s post) and others are telling. The most recent report from Goldman Sachs notes that the stock returns this year for Nvidia are about 165% while they are only about 25% for the infrastructure suppliers (Alphabet, Microsoft, Amazon) and about 1% on average for AI software and the final users (the latter being everyone else). Many of my posts have been talking about the small AI software revenues and productivity growth for final users. Without #productivity growth, why should companies be spending money on infrastructure services to do their AI experiments? The article’s details are even more interesting. The stock prices for Nvidia and the infrastructure suppliers are now trending about 15% above its “forward earnings per share” after being in line with earnings per share for years. The stock prices for AI users (everybody else) have fallen below the trend line. The obvious solution is to move money from Nvida and infrastructure suppliers to the other stocks. Apparently, the stock market believes that all the benefits for AI will go to the suppliers and none to the users. But then why are the users experimenting with AI and purchasing services from the infrastructure suppliers if no one expects them to benefit from AI? It’s almost as if everyone knows there are few benefits from AI, but that Nvidia and the infrastructure suppliers will make big money until the users figure that out. On the other hand, the infrastructure suppliers are profitable unlike those telecom suppliers of the dotcom bubble. The infrastructure suppliers are selling their cloud services for a good price; the question is how long their customers will keep buying the services. A related story from David Cahn of Sequoia Capital is that many of the infrastructure cloud suppliers such as Google (spending $48bn this year) and Meta (spending $35 to $40bn this year) are stockpiling Nvidia chips. He says that the idea “companies need to shell out cash to stockpile computing chips is a delusion that has “spread from Silicon Valley to the rest of the country, and indeed the world.” We know there are some success stories. Advertising videos and call centers seem to represent growing applications for generative AI and then there are the e-commerce, search, and news applications that have been the target for algorithms, Big Data, and now #AI for many years. But for the latter, it has been hard to distinguish between those three, and for the expensive experiments to work out and thus justify the big share prices, bigger applications must succeed. And apparently, Citadel’s Ken Griffen, Ed Yardeni, Baidu’s Robin Li Yanhong, Sequoia’s David Cahn, and other noteworthy investment leaders don’t think they will emerge soon. #technology #innovation #artificialintelligence #startups https://lnkd.in/gubPFnYv

592

145 Kommentare

I analyzed 100+ AI-powered sales funnels. Here's why 95% fail. Most B2B companies waste their AI investment. After 3 months of research, I found the 5 big mistakes causing AI sales funnels to fail: 1. Bad data → feeding poor data into AI leads to bad results. 2. No human touch → removing the human element is a big mistake. 3. Too much automation → automating everything makes customers notice. 4. No personalization → using generic AI responses is a huge turn-off. 5. Ignoring analytics → setting it and forgetting it means no results. But here’s the thing. The top 5% are seeing 5-10x ROI on their AI funnels. How? They mastered the AI + Human hybrid approach. I’ve put together a step-by-step playbook from the top 1%: → Data cleaning tactics → AI-Human collaboration framework → Hyper-personalization strategies → Continuous optimization techniques Want the full playbook? Comment "AI Funnel" below, and I’ll send you. P.S. Using just ONE of these strategies increased a client’s conversion rate by 37% in 30 days. Imagine what all five could do for your business.

8

10 Kommentare

🚨 [AI RESEARCH] The paper "AI, Algorithms, and Awful Humans" by Daniel Solove & Hideyuki MATSUMI is an ESSENTIAL read to understand the debate on 🚧 AI risk management & human oversight 🚧. Important quotes: "Some might be tempted to equate thinking to rationality, with emotions clouding lucid thought, but there are many dimensions to human decision-making beyond rationality, and these nonrational elements are often underappreciated. Professor Martha Nussbaum aptly argues that “emotions are suffused with intelligence and discernment” and involve “an awareness of value or importance.”40 Emotions are “part and parcel of the system of ethical reasoning.” Yet algorithms do not experience emotions. Algorithms might be able to mimic what people might say or do, but they do not understand emotions or feel emotions, and it is questionable how well algorithms will be able to incorporate emotion into their output." (p. 1930) - "The hope that humans and machines can decide better together is not just vague and unsubstantiated; in fact, strong evidence demonstrates that there are significant problems with combining humans and machines in making decisions. Humans can perform poorly when using algorithmic output because of certain biases and flaws in human decision-making.59 Far from serving to augment or correct human decision-making, algorithms can exacerbate existing weaknesses in human thinking, making the decisions worse rather than better. As Professors Rebecca Crootof, Margot Kaminski, and Nicholson Price observe, a “hybrid system” consisting of humans and machines could “all too easily foster the worst of both worlds, where human slowness roadblocks algorithmic speed, human bias undermines algorithmic consistency, or algorithmic speed and inflexibility impair humans’ ability to make informed, contextual decisions." (p. 1935) - "Good qualities in decision-making include a commitment to the scientific method, humility, feedback loops, fairness, morality, lack of bias, empathy, due process, listening to all stakeholders, diversity, practicality, accuracy as to facts, critical reflection, philosophical depth, open-mindedness, awareness of context, and much more. Some decisions might call for more accuracy, but others less so. For landing a plane, we want high accuracy, but for decisions about school admissions, credit scoring, or criminal sentencing, other values are also quite important. There is no one-size-fits-all approach to regulating AI, as the decisions it will be employed to help make are quite different and demand different considerations." (p. 1939) ➡️ Read the paper below. 🏛️ STAY UP TO DATE. AI governance is moving fast: join 36,400+ people who subscribe to my weekly newsletter on AI policy, compliance & regulation (link below). #AI #AIGovernance #HumanOversight #AIRegulation #AIRiskManagement

195

14 Kommentare

Are you interested in the nexus between #ai and #decentralized technology? Didn't make it to #ethcc and want to get the Alpha on the best projects and upcoming events in the space? For example two free tickets to #Bali for the ICP AI Chain Fusion event? AI is great. But when does the hype translate into revenue? How can emerging challengers compete against giants like #OpenAI? What are the most exciting AI subsectors right now? ​This event in #Brussels was amazing. We had a panel of amazing speakers talking shop: ​Jarrod Barnes, Head of Ecosystem & Partner Development, Near ​Emilio Canessa, Head of Global Adoption, DFINITY (ICP) Lou Kerner, Head of CryptoOracle Labs AI Web3 Accelerator Player1Taco .eth, AI & Decentralized Systems Expert Posting this today instead of Monday as a Happy Bday to Kaitlin Argeaux who founded CryptoMondays London in 2018 and who is seen in the photo with four other CryptoMondays leaders in the highlight video including me, Lou, Irina Konova John Goldschmidt. @nav Real leaders like Kaitlin are consistent, reliable, dedicated and know how to treat people. She volunteers her time and energy to make the best events possible and participate in others' activities, always bringing insight, value and fun to whatever she does in the space. Highlight video of the event at the link, please susbscribe! CryptoMondays Global is perfect short video content for people interested in broad themes and learning about #Web3. If you are a long-term investor in technology and want some jokes - also best ideas and projects - subscribe and hear new stuff from leaders across public, private, and the world. If you want to sponsor or partner somehow get in touch. And we would love to see you in one of our 57 cities doing the meetups. Brussels Rocked!!! https://lnkd.in/dx7kPGbH ​Moderated by Ksenia Drobyshevskaya, lead of CryptoMondays Paris. Unfortunately due to video issues the highlight video doesn't include her amazing story of how she took over from Michael Amar 🎗️who then founded Paris Blockchain Week Summit. Since then she's quadrupled the #Paris community size in the last two years and there like all CryptoMondays we have a beautiful range of meetups and keynote events on the industry's most important topics. And many thanks to our sponsors: CryptoOracle Collective https://ethswarm.org and DFINITY. Attend CryptoMondays for IRL love and subscribe to the Myco channel to get the highlights, interviews and Alpha! https://lnkd.in/dKEw-6CH

12

🌟 How to Extend Your Runway by Sequoia Capital 🌟 This insightful document covers three main areas: 1️⃣ What is your runway right now? How should you calculate it? 2️⃣ How should you think about how much runway you actually need? 3️⃣ How do you actually extend your runway if you need more? ~~~~~ Credit: Sequoia Capital, Fazlur Shah ♻️Repost to inspire your Startup Network💡 LinkedIn: Dr. Michael Tadros Agile | PMP® | cPgM® | BSc. | MSc. | PhD. #startup #startups #michael_theprofessor #startupfunding #success #research #partnerships #fundraising #investing #entrepreneurlife #innovation #founders #projectmanagement #smallbusiness #entrepreneurship #entrepreneur #entrepreneurs #venturecapital #investment #venturecapital #siliconvalley #technology

7