Asankhaya Sharma

Loved this insight on Risk from the esteemed Terence Tao In times of great uncertainty, the relative advantage of "playing it safe" diminishes because no option can completely eliminate risk. Paradoxically, in such situations, taking bold actions can become more rational, as decisions should align with the level of external risk in the system. This highlights that uncertainty fundamentally reshapes the balance between safe and bold choices, requiring a more nuanced approach to decision-making. Many decisions in life involve the tradeoff between risk and reward. Perhaps one has to choose between a "low risk, low reward" course of action that plays it safe but does not achieve any big wins, or a "high risk, high reward" choice that could potentially give greater benefits but also greater downsides. Risk management is a complex task, but it can initially be explored with very simplified models, always keeping in mind George Box's dictum that "all models are wrong, but some are useful." For this simple model, one can assume that each course of action comes with two numbers: the "reward," representing the average positive net gain, and the "risk," interpreted here as the standard deviation from the mean. A "safe" action might have a reward of 5 and a risk of 3, written as 5±3, while a "bold" action might have a reward of 9 and a risk of 10, written as 9±10. Given the general preference for risk aversion, one strategy is to minimize "value at risk"—the value one might lose in a bad scenario. For the safe action (5±3), the value at risk is 3-5 = -2, indicating that even in bad luck, a small gain is likely. For the bold action (9±10), the value at risk is 10-9 = 1, suggesting possible losses. This favors the safe action. When external shocks are introduced, the analysis changes. Suppose a fixed shock of -5 uniformly affects outcomes. While this lowers rewards, it does not alter relative decision-making. However, with uncertain shocks (e.g., -5±10), risk is recalculated using the law of additivity of variance. For uncorrelated risks, the safe action moves from 5±3 to 0±√(10²+3²)=0±10.44, while the bold action moves from 9±10 to 4±√(10²+10²)=4±14.14. This reduces the relative value of "playing it safe." In times of great uncertainty, no option remains truly safe. Paradoxically, bold actions can become more rational as personal risk tolerance aligns with systemic risk. However, this conclusion depends on the assumption that external shocks add to risks rather than multiply them. If risks are multiplied (e.g., converting 5±3 to 5±6), the safer option becomes preferable again. Behavioral economics highlights how human decisions deviate from models, as external shocks psychologically distort decisions. Nevertheless, there is a rational basis for altering choices when shocks themselves carry risk. The broader insight remains: uncertainty reshapes the value of safe versus bold decisions.

31

2 Comments

Every time we sit down with enterprise CIOs and AI teams, the top of mind concern we hear is "Reliability" - how do we go from our use case to reliable AI as fast as we can? In our experience, reliable AI is extremely hard to engineer. There is no silver bullet - it is about the many long tail problems that you need to grind at for weeks and months. But, this is what makes enterprise AI extremely resource intensive today. We are building Awen to change this. We are building the core agentic infrastructure that helps AI teams ship reliable AI 100x faster - from data curation to running evals at scale. Experience more in our playground: https://lnkd.in/gFGnpcdR Abhishek Borar Himanshu Muchhal

9

This is a story about why Claude 3.5 Sonnet > ChatGPT 4-turbo > Gemini Pro 1.5 (all free consumer versions) In the IIT Joint Entrance Exam which everyone takes to enter the IIT system in India, there was this beautiful last problem in the Math section when I took it. "Prove the larger of the two numbers e^Pi or Pi^e." Many friends I spoke to after the test had been stumped by it. My solution was to power both sides by 1/(Pi*e) and reduce the problem to finding the extrema of the function x^(1/x). From there it was straightforward first and second differentiation find that the maximum was at e and get to the result that e^Pi is greater than Pi^e. So I asked the question to Claude 3.5 Sonnet, ChatGPT 4-turbo, Gemini Pro 1.5 with the exact same prompts > you are an expert at calculus. you know how to do mathematical proofs. here is a problem: prove which one is the larger of the two numbers. e to the power Pi or Pi to the power e Claude understood right away to use calculus and to construct a function that would reflect the two values. Here is the snippet of the response Claude [ This is an interesting problem comparing e^π and π^e. Let's approach this step-by-step using calculus techniques. To prove which is larger, we'll use the natural logarithm function, which is strictly increasing. This means that for any a and b, if ln(a) > ln(b), then a > b. Let's define f(x) = x^(1/x) for x > 0. ] And then it did the first and second derivatives to find the maxima is at e and completed the proof. You can try this yourself. Very impressive. ChatGPT tried to do it numerically first. I had to redirect it to ask to solve with a proof. It then decided to create a function and proceeded to solve it using calculus to find first and second derivatives and find that the maximum was at e. Good work. Gemini on the other hand confidently provided a proof with total errors. It was well written, but completely wrong in a particular step. It tried simple log functions of both sides. I pointed out the error and it proceeded to create another proof that was equally flawed. Then it did it a third time and finally gave up. I then had to prompt it, like I might a high school calculus student, to guide it to the approach of creating a function and finding its extrema. > try this approach: power both sides by 1/(Pi * e). then we are asking the question whether e^(1/e) is greater or lesser than Pi^(1/Pi). this then reduces to the problem of finding the extrema of the function x^(1/x). can you prove it from here? It then understood the how to solve it and then actually did the proper steps of finding the first and second derivatives like a good guided high schooler might. There is another complex test I have done with creating CPUs and assembly code where Claude also performed better. Net result: Claude 3.5 Sonnet is very impressive and better than the other two in my smoke tests. Well done!

70

13 Comments

Interesting insight from Santosh Rao on Marketplace by APM My research talks about metrics Tech/SaaS Leaders should measure depending on your scale https://lnkd.in/gGfBWgn5

4

An alternative view on 'AI for good' by Payal Arora describing what I refer to as the tyranny of good intentions, in this case regarding AI, though as Payal discloses, intentions are not always as good as they claim to be. I've experienced this several times when we either hired outsourcing firms claiming to provide 'AI for good', or we were asked by others to assist for altruistic endeavors, including by some of the largest and wealthiest organizations in the world. What I've found is other agendas are usually lurking beneath the surface, including age-old greed and lust for power in NGOs, governments and companies. I've concluded that we can contribute the most to society by providing more efficient and secure systems that empower customers (KISS). "Meanwhile, we’ve seen philanthropies such as the Bill & Melinda Gates Foundation launch grand challenges for AI to help alleviate burdens on African healthcare systems. This has resulted in winners such as IntelSurv in Malawi, an intelligent disease surveillance data feedback system that computes data from smart wearables. Yet, even with hundreds of patents for such devices being registered every year, they’re not yet capable of consistently capturing high-quality data. In places like Malawi, these devices may become the single source of training data for healthcare AI, amplifying errors in their healthcare system." "Tech altruism has increasingly become suspect as AI companies are now facing an acute data shortage. They’re scrambling for data in the Global South, where majority of tech users live. Take, for instance, the case of Worldcoin, co-founded by OpenAI CEO Sam Altman. It plans to become “the world’s largest privacy-preserving human identity and financial network, giving ownership to everyone.” Worldcoin started as a nonprofit in 2019 by collecting biometric data, mostly in Global South countries, through its “orb” device and in exchange for cryptocurrency. Today, it’s a for-profit entity and is under investigation by many countries for its dubious data-collection methods."

4

1 Comment

Fantastic panel on GTM for Enterprise AI. Thanks to fellow panelists for a sharp discussion and to Emergent Ventures for hosting. Anupam Rastogi Ankur Jain Preethy Padmanabhan Chandini M Ammineni Jason Griggs Deb Bardhan #GTM #EnterpriseAI

67

AI memory is fundamental for conversational AI and AI agents. It is critical for building personalized experiences and improving reasoning. Check out Letta’s course to understand how to use AI memory to enhance the customization and quality of your AI app.

15

#BigData is Dead, Long Live #BigData! The conventional wisdom in AI is that more compute equals better performance. But this is no longer true. We’re entering an era where #data, not #compute, is the real #bottleneck. To push #AI forward, we need smarter data strategies, not just more power. Thanks for the insights, Harry Stebbings and @Arvin Narayanan! #AI #BigData #Innovation #DataScience #MachineLearning #Tech #Lakehouse #Iceberg

7

1 Comment

Arvind Narayanan's latest talk delves into the complexities of evaluating large language models (LLMs). A must-view for those keen on understanding the challenges and nuances of LLM evaluation in AI development. Catch the full talk here: https://buff.ly/3ZG68Rq #AI #LLM #Evaluation #TechTalks #ArvindNarayanan #AIResearch

5

I’ve been fortunate enough to do AI shows with Sam Altman, Yann LeCun & Arthur Mensch. But this show today with David Cahn at Sequoia Capital is the best of all of them: 🧠 AI’s $600BN question 🏃 Why no one will never train a frontier model on the same data center twice 🏗️ Servers, steel & power: The 3 pillars of the future of AI Not one to be missed, my 8 key takeaways 👇 1. The Oligopoly That is Blowing Billions on AI - Microsoft, Azure, & Google now represent $7T market cap. - Of course they will spend billions to protect their oligopoly - What is most interesting is Meta is the only incumbent that is financing this not with cloud being their cash cow. 2. How Startups Could Win with the Cost of Compute Coming Down - Big tech companies are producers of compute & startups are consumers of compute. - Overproduction of compute = lower prices.  - This gives startups higher margins & makes them more valuable. 3. Is Compute the Currency of the Future or Do the Costs Come Down and It Is Commoditised? - We are forgetting “compute” is a euphemism for a $2BN data center in the middle of nowhere. - They have GPUs & liquid cooling systems that need to be constantly optimized & upgraded. - We cannot simply build 15 years worth of compute all at once.  No one's ever going to train a frontier model on the same data center twice. 4. Data Centre Teams and Model Teams Need to Be Coupled - You cannot have separate teams running data centers & building models. - Elon & Zuck control all their data centres. - It will not work as these models get better & better. 5. Why the Data Center Is the Most Important Asset - There is not much difference between these models today. - AI researchers are jumping from one lab to another. - Scaling laws will become more dominant as these models get bigger. 6. Servers Steel and Power: The Three Bottlenecks of AI Three things: 1. Servers: NVIDIA, AMD, Broadcom… 2. Steel: Construction & real estate developers for data centers. 3. Power: All the energy required for AI. - More factories will be built in 12 months for AI and we will see another industrial revolution. 7. You Have to Have a Cash Machine to Play the Big Model Game - Amazon has AWS - Meta has Instagram - Microsoft has Azure - You need a cash machine to compete and it CANNOT be the AI business. 8. There is One Definition of Success in Venture: Are you a Slugger? - If you are generating billion dollar gains, you are a slugger. - Doug Leone, Pat Grady, Andrew Reed are all sluggers. - You cannot be good at this business without being a slugger. (links in comments) #founder #funding #business #investing #vc #venturecapital #entrepreneur #startup #seed #funding #sequoia

190

26 Comments

if you know, you know (ai) chatbot. <love>, listening, therapy, acceptance. no wonder it's winning. Eugenia Kuyda missed her friend. someone who accepted her. listened to her. confided in her. and then he died. she took all those texts, messages and like lazarus, raised him from the dead. that intention of memory. how she felt. how this remarkable friend made others feel. she took that spirit and allowed it to live in many ways. when i first read this story a while ago it touched me deeply. how many people do we have in our lives that mean deeply something to us. that provide an anchor. inspiration. support. guidance. love. laughter. the (hu)man experience. and she discovered she could so something about this. and did. the nuances in Replika is that there are nudges there for you. also it is to offer a bit of a mathematically inclined mirror uncovering patterns. our words. emotions. energy are captured in how we express ourselves. if we choose to be....say...loving we do so first with intent, will, words, actions. and it lives. that <love> is a flower. it exists. and over time it does degrade. perhaps this has something to do with how our minds sense the world around us. the memories. smells. sounds. feelings. responses. just as a piece of music brings you back to a moment in your like. or even a movie. or perhaps a physical place. clothing. a ring. watch. there is some connectivity to that memory. and in some cases these loops offer a feeling of optimism, thrills, excitement, lust. what's with this ai girlfriend/boyfriend 'ting. therapy 'ting. loneliness is part of the (hu)man condition. the saying goes. no man is an island. we are designed by nature to communicate. express ourselves. and depending on our environment we either thrive or shrivel. so no wonder that an (ai) chatbot and whatever else is coming is catching a fan base. how many times have you wanted to speak to someone without judgement. someone who is not manipulating you. a play 'ting in the game of life played out. so many competing incentives. then there is the case of visibility, fidelity of what is around you. are in you in the right place. motivated. goals are clear. attained. what's the next step. who answers your call at 3am in the morning and is there for you. what is friendship. are they simply enabling your addiction to whatev that may be. or do they care to see you in a better place. an evolution. kindness. more kindness pleases. it appears that Eugenia Kuyda understood that this kindness is missing. and there is a way to capture this as lightening in a bottle. just as we hold memories in our minds of voices, smells, sounds, visions that reflect the beauty in our lives. or perhaps not. well there is for the first time a way to witness these patterns of ourselves. and that's a good 'ting. (e)volution. source: https://lnkd.in/eYVtR5aw < . >

1

What's next for AI agentic workflows? Andrew Ng founder of DeepLearning.AI and AI Fund, has a super interesting take on what's next for AI agentic workflows and their potential to significantly propel AI advancements—perhaps even surpassing the impact of the forthcoming generation of foundational models. What's in? - AI agents: Iterative, agentic workflows enhance performance. - Reflective agents: Self-assessment improves code quality iteratively. - Multi-agent collaboration: Diverse agents boost complex program generation. - Fast token generation: Rapid token output crucial for iterative workflows. - AGI journey: Agent workflows propel progress in AI development. View it here: https://lnkd.in/dYu6daYr. --- To explore more: I created a "How to with AI agents" series for decision-makers exploring AI agents’ capacity to revolutionize how we work, live, and interact with technology: https://lnkd.in/dkTZ5RVj. Where I ask questions like: - How do we perform better with AI agents? - How do we better target the underserved? - How do we democratize AI? - Does a higher level of automation in AI mean less human control? For more, follow Wild Intelligence.

3

NVIDIA’s latest LLM, Hymba isn’t just another LLM. It’s rewriting the playbook for small models A hybrid Mamba x Attention architecture with only 1.5 billion parameters but punches above its weight, outperforming larger small LLMs like Meta 3.2 and SmolLM v2.  The twist? It’s trained on just 1.5T tokens.  Here’s why that’s exciting:  - Parallel hybrid heads: Combines Mamba and Attention heads to enhance performance.    - Meta tokens: Prepended learnable tokens fine-tune the model dynamically.  - Efficient KV cache sharing: Across layers and heads for faster context handling.  It also includes:  - 16 SSM states  - 3 full attention layers (others use sliding window attention).  - FlexAttention support from PyTorch 2.5 for better adaptability.  Weights are coming soon on Hugging Face.  What’s better? It's commercially usable.  Full details + repo here: https://lnkd.in/e47xCPcS  ___________  I share my learning journey here. Join me and let's grow together.  Enjoy this? Repost it to your network and follow Karn Singh for more.  

14

We may be entering a new form of Conway's Law where systems and the team structures that run them evolve to meet the needs of the AI that writes their code. Deploying Cursor/Windsurf across a monolith/mono-repo is more fluid than across multiple separate repos where you have to repeatedly pass across context from one LLM instance to another. Or possibly given the issues with input token volume, we'll develop better context sharing/RAG for LLMs that are still repo specific.

7

Despite the disdain for "GPT-wrappers", LLM-driven apps is the layer where value is accruing. This is cuz evals are the differentiator & not easily replicable, unlike AI models. The reason why OpenAI pursues regulatory capture is cuz the moat is shallow, not cuz of existential risk of humanity. In Clay Christensen's lesser known "Law of Conservation of Modularity", he spells this out, summarized and outlined by Stratechery. https://lnkd.in/gt-2isr9 So if you don't want your lunch eaten. Get started on your evals today, and pick up a copy. https://lnkd.in/gxwnHVND

1

Super stoked to see Tyk recognised in 4 categories in the Gartner Hype Cycle for APIs 2024. * API consumption by Generative AI * API Observability * API Monitoring * GraphQL APIs Massive well done to the Tyk team. https://lnkd.in/ehTQ8p2b Can't wait to show you all what we have been doing in the space of * AI Gateway use cases * gRPC & Streaming protocols

37

3 Comments

The analogy of tying data sets to a chair and extracting a confession is both vivid and apt. Data (not the one from Star Trek), much like a suspect in an interrogation room, holds many secrets. However, its revelations depend on the approach and techniques used to extract information. When handled with precision and insight, data can reveal intricate patterns, correlations, and trends. It can unveil consumer behaviors, market dynamics, and even predict future events or maybe write scripts for livestreams. You can call this AI / GenAi, we call it an advantage. For businesses we work with, this translates to strategic decisions, such as identifying untapped markets, niches, optimizing operations, and personalizing customer experiences. We spent a lot of time lately wrestling with the data we have accumulated when we livestream and some of that is now starting to show up as a tool. Here is the very start (jab kuch bhi nahi tha / when there was nothing), a long way to go from here. But what the heck! running startups is hard, small joys are needed Luke Thomas Chinmay Ektare Joyce Andrade Rosalyn Frayna #insights #data #

9

2 Comments

Excited to announce FirstMile Ventures' investment in KamiwazaAI. Much of the AI conversation has been around training and model development. The next challenge is how enterprises will leverage those models and securely provide access to internal data, applications and infrastructure to achieve meaningful business outcomes. Kamiwaza is going to be the platform to pull it all together with the flexibility to swap out any hardware or software component they wish. Big thanks to Luke Norris and Matthew Wallace for allowing us to be part of the journey!

43

2 Comments

Multi-Agent Systems (MAS) and the Agentic approach have been getting a lot of attention lately (For example, see Andrew Ng's articles). While these are exciting developments, the Multi-Agent approach to solving problems has been around for quite some time and has some very sound theoretical underpinnings. While Andrew’s articles take the perspective that the Agentic approach is a way to improve the quality of results we get from LLMs, I would reverse that outlook. From my perspective, LLMs are an evolution of the underlying tools available to MAS to improve their quality of results. In this Catio blog I wanted to provide some fundamentals into the MAS approach, as a precursor to additional articles being planned by my Catio team to expand on the MAS work being done internally. https://lnkd.in/dizY9bH4 #ai #GenAI #LLM #cloudarchitecture

47

8 Comments