Ehsan Khan

AI in Finance: From Academia to Groundbreaking Projects 💲📈    In this episode of #BehindtheGrowth, I am celebrating  Zain Abbas, Ph.D., AI and machine learning expert, and Director of Data Science at Sun Life.     Sitting with Zain gave me the opportunity to learn about his transition from academia to the industry, gain insights into the impacts of generative AI and future trends, and explore the use of predictive analysis for financial services.     Whether you are a seasoned professional or just starting out in the finance industry, tune in to the full conversation to gain valuable insights into the challenges and solutions of career change, and strategies that drive career growth.   Find it using the links below: 🔗     Website: https://lnkd.in/gCxx8mMV      YouTube: https://lnkd.in/gEh9C-tY      Apple: https://lnkd.in/gVDngQSA      Spotify: https://lnkd.in/g5G9T-RC     @mobileLIVE #Finance #CareerGrowth #CareerChange #GenAI #PredictiveAnalysis #AI #Development #MachineLearning  

239

10 Comments

Researchers evaluated GPT-4 and ChatGPT in USMLE soft skills, assessing their performance in judgment, ethics, and empathy. GPT-4 outperformed ChatGPT, answering 90% of questions accurately compared to ChatGPT's 62.5%. GPT-4 never revised its responses, showing confidence, while ChatGPT revised 82.5% of answers. Both models demonstrated proficiency in soft skills, with GPT-4 excelling in professionalism, ethics, and empathy. The study suggests AI models can effectively handle tasks requiring human-like skills.

2

2 Comments

It was a superb chatting with Andrew Beam and Arjun Manrai. Thank you both for the long form conversation! Refreshing break from the "bite sized" world enabled by NEJM AI. What we talked about was possible given the stellar research team I get to work with (Jason Fries, Alison Callahan, Michael Wornow, Suhana Bedi, Akshay Swaminathan, Zepeng Frazier Huo, Scott Fleming and Ethan Steinberg) and the vibrant community at Stanford University School of Medicine, Stanford Health Care supported by the brilliant team at Technology & Digital Solutions - Stanford Medicine. Some of the ideas we discussed are also outlined in blogs at our Stanford Institute for Human-Centered Artificial Intelligence (HAI). e.g. 1. How can foundation models advance healthcare - https://lnkd.in/eiBw4YHr 2. How do we ensure healthcare ai is useful - https://lnkd.in/eDTc7Dqi 3. Are we there yet with LLMs in healthcare? - https://lnkd.in/eDQv2zH4 #HealthcareAI

277

4 Comments

👉🏼 Investigating influencing factors of learning satisfaction in AI ChatGPT for research: University students perspective 🤓 Mohammed Abdullatif Almulla 👇🏻 https://lnkd.in/eYu7c93c 🔍 Focus on data insights: - Interaction learning and collaborative learning are pivotal in fostering ChatGPT adoption. - Information quality substantially influences researchers' decisions to continue using ChatGPT. - Perceived ease of use and perceived usefulness play intermediary roles in linking ChatGPT engagement to learning satisfaction. 💡 Main outcomes and implications: - User-friendly interfaces and perceived utility are crucial factors affecting overall satisfaction levels. - ChatGPT positively impacts learning motivation, enhancing student engagement and interest in learning. 📚 Field significance: - Educational practitioners can improve the implementation of AI technologies in university students by focusing on user-friendly design and collaborative learning. - Continued research is needed to comprehensively understand the interplay between AI-powered tools, learning objectives, and motivation. 🗄️: [#ChatGPT #learning satisfaction #AI adoption #interaction learning #collaborative learning]

3

I recently participated in a fascinating meeting at Tuwaiq Academy titled "ثقافة التحليل في الذكاء الاصطناعي التوليدي و النماذج اللغوية الكبيرة (GenAi & LLM)", Professor Alan Jacobson led the discussion offering valuable insights into these technologies... Professor Jacobson's expertise and enthusiasm were truly inspiring. His guidance through the complexities of Generative AI and LLMs sparked insightful conversations and a deeper understanding of their potential impact... The meeting highlighted the transformative power of Generative AI and LLMs across various industries, from healthcare and education to finance and beyond. I'm excited to see how these technologies will continue to shape the future. #GenerativeAI #LLMs #TuwaiqAcademy #AI

11

👉🏼 How effective are large language models in product risk assessment? 🤓 Zachary A Collier 👇🏻 https://lnkd.in/eZy9nVrh 🔍 Focus on data insights: - Large language models, like ChatGPT, show promise in brainstorming potential failure modes and risk mitigations. - However, there were errors and inconsistencies in some results. - The guidance provided by the models was perceived as overly generic and occasionally outlandish. 💡 Main outcomes and implications: - ChatGPT performed better at divergent thinking tasks but lacked depth of knowledge compared to subject matter experts. - Similar patterns in strengths and weaknesses were observed when comparing ChatGPT to other large language models. - Despite challenges, there may be a role for large language models in assisting with ideation in product risk assessment. 📚 Field significance: - Product safety professionals can leverage large language models for brainstorming and idea generation in risk assessment processes. - Experts should focus on critically reviewing AI-generated content to ensure accuracy and relevance. 🗄️: [#productrisks #AI #riskassessment #ChatGPT]

1

1 Comment

Simplify Multi-Agent Frameworks There is a new AI framework called Llama-Agents; it's open-source and makes building, managing, and deploying multiple AI agents easier. The agents work as independent microservices, which helps with scalability. They all communicate through a central control system that assigns tasks efficiently. Its flexible setup lets you define specific task flows or use an orchestrator to manage them dynamically. Plus, deploying and scaling these agents is straightforward, making it great for small and large projects. Check it out https://lnkd.in/ev9_mpnB.

16

1 Comment

2024 AI INDEX REPORT ⬇ A detailed analysis of recent developments in artificial intelligence, emphasizing both the advancements and the challenges faced in the field. The report offers an extensive overview of the artificial intelligence landscape, highlighting rapid advancements, increased global investment in AI, and significant shifts in public opinion. 1️⃣ AI's capabilities are expanding, with notable improvements in language understanding and object recognition, although complex tasks like common sense reasoning remain challenging. 2️⃣ Investment in AI, especially in generative AI technologies, has surged, with significant increases in both private funding and public sector engagement. 3️⃣ Public perception of AI is mixed, with growing awareness but also increased concerns about privacy, misinformation, and job impacts. 4️⃣ Policy and regulatory landscapes are evolving rapidly to keep pace with technological advancements and address societal concerns. 5️⃣ Advances in AI are significantly influencing industries and domains such as healthcare, where AI-driven tools and systems are increasingly adopted. ✍🏻 Nestor Maslej, Loredana Fattorini, PhD, Dr. Elif Kiesow Cortez, Anka Reuel, Robi Rahman, Alexandra Rome, Lapo Santarlasci, James da Costa, Simba Jonga, Curtis Langlotz, Russ Altman, Jack Clark, Raymond Perrault, Erik Brynjolfsson, John Etchemendy, Katrina Ligett, Terah Lyons, James Manyika, Juan Carlos Niebles, Vanessa Parli, Yoav Shoham, Russell Wald. The AI Index 2024 Annual Report. Stanford Institute for Human-Centered Artificial Intelligence (HAI). April 2024. 💥 Join our community of 30,000 curious minds! ✅ Subscribe to our newsletter for your front-row seat to the latest groundbreaking studies. Get started here: https://lnkd.in/eR7qichj.

148

3 Comments

THE ADOPTION AND IMPACT OF GENERATIVE AI IN VARIOUS INDUSTRIES ⬇ This report provides a comprehensive analysis of the adoption and impact of generative AI in various industries, detailing the current challenges and future directions in AI implementation. 1️⃣ The survey includes insights from over 1800 ML practitioners and leaders, focusing on their experiences with AI, emphasizing a significant gap in AI safety and security benchmarks. 2️⃣ Adoption trends show a rise in closed-source AI model usage driven by cloud service partnerships, while interest in open-source models persists due to their flexibility. 72% of organizations aim to increase investment in commercial closed-source models, with 67% planning to invest in open-source models. 3️⃣ A significant 61% of organizations cite that insufficient tooling is a major barrier to AI adoption, alongside other challenges such as budget constraints and data privacy issues. 4️⃣ The report highlights a shift towards more sophisticated evaluation methods and the growing importance of domain-specific data and fine-tuning to enhance AI model performance. 5️⃣ The survey indicates a substantial shift towards the adoption of generative AI, with a significant rise in the use of closed-source models from 37% in the previous year to 86%. 6️⃣ Only about half of the organizations measure the business impact of their AI initiatives effectively. ✍🏻 Scale Zeitgeist: 2024 AI Readiness Report. March 2024.

36

👉🏼 The ChatGPT: Revolutionizing Research with AI 🤓 Brítez Mirta 👇🏻 https://lnkd.in/eqRCkXaF 🔍 Focus on data insights: - Utilization of ChatGPT for data analysis - Enhanced data interpretation through AI - Improved decision-making based on data-driven insights 💡 Main outcomes and implications: - Accelerated research processes - Increased accuracy in data analysis - Enhanced understanding of complex datasets 📚 Field significance: - Advancement in research methodologies - Potential for transformative impact in various fields - Integration of AI for improved research outcomes 🗄️: [#ArtificialIntelligence #Research #DataInsights #AI #ChatGPT]

2

“Big consulting firms are doing everything from developing plans for AI customer service bots for some clients to streamlining R&D for healthcare companies. Some also track and make sense of the rapidly changing laws and policies around the tech. McKinsey said generative AI will make up about 40% of its business this year. IBM’s 160,000 consultants have locked in $1 billion in generative AI sales commitments. Consulting firm BCG’s CEO Christoph Schweizer told the Financial Times, “We have never seen a topic become relevant as rapidly as Gen AI.”” @morningbrew If you’re interested in exploring how AI can help your company or if you have a specific implementation or solution to be built, our team @bitbakeryco can help.

2

1 Comment

On May 3 2024, OpenAI posted a blogpost on their recommendations for securing infrastructure to avoid AI misuse. This is nearly a year after Sam Altman proposed having AI regulation setup and a separate AI regulation agency during his hearing on 16 May 2023. Regulation has been the tool used by incumbents to create a massive entry barrier for new players. It has never worked for the common good, especially the new entrants. OpenAI is going the same route as the other industry counterparts like pharma, healthcare, energy, etc., which was historically avoided by the silicon valley. And it is worrying. While their long blogpost should be read by every AI practitioner and developer (link in comments), one highlight from it that is a definite NO is - 'cryptographically attesting GPUs'. It means, GPUs should be certified to run AI models (more in the context of LLMs, AGI & Super Intelligence). Else, you'll not be able to run / build AI models on them. Now, imagine its implications. 1. A start-up needs these regulated, certified & a lot more expensive GPUs to use existing AI models. 2. A new chip company needs to go through the process of getting them certified for end-users to consume them. 3. A new AI model must be certified to run on these certified GPUs. In other words, the model weights could get intertwined with these GPU certifications! These are unnecessary. They stifle the core of technology philosophy which guided open innovation for the past 3+ decades. Their other blogpost on AI Agent behaviour had decent coverage with important & influential people/teams talking and discussing about it. Was surprised that this did not get the coverage & dissent it deserves. #AI #Regulation

13

1 Comment

As consultants integrate LLMs into their mainstream workflow to solve problems internally and for clients, it should accelerate the use of LLM tools to reduce administrative costs amounting to 25% of healthcare expenditure and over 500 billion dollars per year. This is over triple that of our closest industrialized nation. Much of this is the nature of our healthcare system, but will AI make a dent in this expenditure or will one cost replace another?

2

1 Comment

💲 "Administrative AI" vs "Clinical AI" in healthcare As several news and data suggest, most of US VC dollars are going towards companies working on AI in healthcare. 1 in 4 dollars invested in healthcare going towards companies leveraging AI! Around ~$7 billion of U.S. VC dollars were invested in AI healthcare companies last year, representing ~20% of total VC healthcare investment. While trends continue to rise, one of interesting area is to see what all these AI companies will be trying to do. This is where things get exciting since administrative side of healthcare is attracting most of attention. Administrative side of healthcare still overwhelming to the industry and consider as 'low hanging fruit' to target first since it has clear ROI with less regulatory hurdles. News around 'Drug Discovery Using AI' may just sound exciting, however faces challenges to get the nods from several parties and not always easy to get passed regulations. To put AI to use towards administrative side of healthcare has several advantages where one can define 'clear execution' and 'transparent usage of investment'. This is one of reason why AI is being tailored to be used for productivity, reduce burden and less manual efforts by staff. Usecases like virtual assistant, clinical notes insights, revenue cycles are just some of scenarios on the rise! Especially in Generative AI space where everyone is still in learning stage! #genai #ushealthcare

11

1 Comment

🌟 The Evolution and Growth of Large Language Models (LLMs) 🌟 Generative AI models, especially LLMs, are reshaping AI by learning from vast datasets to generate new content. 🚀 Here's a quick look: RNNs: Good at sequential data but struggle with long-term dependencies. 📉 LSTM: Improved memory handling but computationally heavy. 🧠 GRU: Simplified LSTM for faster training but still has long-range dependency issues. ⚙️ Attention Mechanisms: Enabled better context understanding, leading to Transformer models. 🎯 Transformers: BERT introduced bidirectional training for better text understanding, while GPT focused on autoregressive language modeling for text generation. 💡 Overcoming Drawbacks: Transformers: Process entire sequences in parallel, enhancing efficiency. 🔄 BERT and GPT: Combined strengths of understanding and generating text, leading to powerful models like GPT-3. 🌐 Applications and Future Directions: LLMs are used in chatbots, translation, creative writing, and coding assistance. Future research focuses on efficiency, bias reduction, and interpretability, promising further advancements. 🌍🔍 This journey highlights the significance and transformative potential of LLMs in modern AI. 💥 #AI #MachineLearning #LLMs #Technology #Innovation #levelupwithgenai

8

7 Comments