Michael Quinn

The best learning during my time at Amazon was improving my writing. Here is my list of how to edit your writing to improve it and how to get an idea down on paper in the first place. TL;DR craft compelling documents by anticipating reader needs, addressing potential questions, and clearly articulating the desired outcome. This will take time and a lot of editing. Make your writing better, editing tips: A. Use data and logic to lead the reader to a conclusion. Using adverb and adjective is less effective. B. Eliminate weasel words C. Use readable scores to get data on how readable your doc is now and improve it D. Use read-a-loud extension or read it out loud yourself E. Edit words that don't add value. Look for repetitive language. Don't tax your reader with big words How to crafting an idea in writing: 1. Start with bullet list of what your thinking about 2. Start by writing with limited edits. Sleep on it and then edit like crazy. 3. Know your reader and write for them. 4. Write down FAQ to the end of a doc that you think your reader will have. If you can, you should answer these in your writing but it's not always possible.  5. Take ownership not ask for ownership. If you are asking for ownership or asking the reader to force someone to do something you are doing it wrong.    6. Ask why 5 times on customer facing questions to get to the detail why this is important 7. Think in 3s. How would you explain this with a 3 word (The Title), 3 sentences (TLDR), 3 paragraphs (a 1 page page overview), 3 pages (longer form with more detail). 8. Bonus: add some long term thinking topics to your doc. If the reader is excited about what you are writing about, give them more and think big about the future. 9. Close with what decision you need help from the reader 10. Finally writing a couple page document should take a week(s) not days or hours to complete.

78

7 Comments

In this article we are providing the major architectural building blocks and blueprint for building end-to-end GenAI-based applications that are ready for production. I’d like to call out several key considerations as you go about implementing and designing these llm-based applications. #llms #llmops #chatgpt4 #aiml #gemini #mlops #data https://lnkd.in/dMev5ncq

8

A new (to me!) way to measure relationships between variables - I have previously worked with Pearson and Kendall but this one is new. This article explores its applications, and I can already think of past projects where it would have been valuable. Intriguing stuff! #ksaai #datanalysis #correlation

59

Is Generative AI just a bubble that's about to pop? Aiden Gomez from Cohere doesn't seem to think so and it's hard to disagree with much of what is said in this interview. https://lnkd.in/gXeJSncZ 1. LLMs are well beyond stochastic parrots - they can apply learned patterns to new inputs. 2. In the vast majority of applications it is about augmenting humans by allowing them to work faster and more effectively. 3. Emphasis is not on making the largest model possible, but on models that enterprises find useful and can reasonably consume. 4. Despite the expensive compute, the path to revenue is not far down the road - but it won't happen by wishful thinking... 5. The real risks are not existential, but rather misapplication/abuse of the technology. 6. The bar set by LLMs is being raised all the time and whilst there may exist an upper limit to their capabilities, it's likely we can get the rate of hallucinations to be at least on par with humans!

4

François Chollet and Mike Knoop have launched a new Abstraction and Reasoning Challenge (https://arcprize.org/) with over $1 million in prizes for those who can surpass the human baseline. I really like this initiative. I participated in the first ARC competition four years ago and took 3rd place with my teammate, Vlad Golubev. This challenge requires developing an algorithm that solves simple logical puzzles—tasks that are easy for humans but difficult for AI. This approach offers a benchmark that is very different from the typical ones for AI that focus on real-world problems, like those used for large language models (LLMs). In previous competitions, many solutions, including mine (https://lnkd.in/gt9G2cz), involved creating a Domain Specific Language (DSL) — an abstract set of possible transformations — and conducting an optimized search through possible combinations of transformations. This and other popular methods were more like "smart brute force" than "artificial intelligence." Subsequent competitions have unveiled new methods that have led to better results. I'm eager to see if any novel approaches in the upcoming contest will deliver a higher level of generalization and intelligence, solving problems in much smarter ways. #AI #AGI #ARC #kaggle #competition #datascience #arcagi

15

Join me at #SnowflakeSummit! I'll be on stage with Ellery Berk sharing new #AdvancedAnalytics capabilities in the #DataCloud. Join this session to explore what’s new in #Geospatial data, #TimeSeries, #search optimization, and #MachineLearning with SQL: https://lnkd.in/g-GjvqWr

99

4 Comments

What people leaders can learn from Large Language Models (LLMs) Because LLMs are trained from human data, it's unsurprising that their behavior mirrors that of their human counterparts. It's worth taking a closer look at human teams through the lens of what we've learned about LLM performance. While it wasn't the authors' goal, this paper highlights many LLM issues and recommendations that parallel what is needed for human business teams. Here are the key takeaways from section 5. - Assigning tasks to individuals based on their specialties improves performance. - A clearly identified leader and a comprehensive feedback system will keep the team on track. - Clear role definition ensures that individuals stay focused and reduces mistakes. - Planning and task decomposition improve decision-making and execution. As a final note, the paper mentions the use of a rotating leader to improve the overall performance of the multi-agent response. Applied to human teams, this is a novel approach with great potential. Leaders who stay connected to technological developments would inject that fresh knowledge into their teams.  What team wouldn't benefit from periodically cycling through people leaders who have kept their skills sharp and relevant? https://lnkd.in/gXUp88th

21

Data aficionados, I've stumbled upon some incredible tools that are real game-changers in our industry. Here’s a quick rundown: 1. Vectorize is designed for Retrieval-Augmented Generation (RAG). It effortlessly converts unstructured data into high-performing vector search indexes. Imagine cutting through data chaos like never before. Do any of you folks already use this? 2. Spreadsheet is all you need—a mind-blowing nanoGPT pipeline neatly packed in a spreadsheet format. Perfect for anyone looking to get a handle on GPT's intricacies without diving into hardcore programming. 3. Replicate makes it a breeze to run and fine-tune open-sourced AI models using a straightforward API. Talk about simplifying complex operations! What AI models are you all experimenting with right now? 4. transformers.js brings the power of transformers right to your browser. Think about the possibilities here—AI execution without ever needing to leave your browser. Have you guys tried this yet? 5. Build your own X is a treasure trove of step-by-step guides for creating your own tech marvels from scratch. Whether it's AR, bots, or torrent tech, this resource has got you covered. Curious, which tech are you excited to recreate? 🚀 It's an exhilarating time to be in the data industry with tools like these at our fingertips. Let's elevate our game! #DataScience #AI #MachineLearning #VectorSearch #GPT #TechInnovation #DIYTech #Transformers Jump in and share your experiences or thoughts about these tools! 🌟

5

1 Comment

Controlling hallucinations produced by LLMs is critical for their effective adoption. This paper is an important step in understanding the dynamics at play. For those deploying LLM this is a very worthwhile read. "As expected, providing the correct retrieved information fixes most model mistakes (94% accuracy)... Similarly, we also find that the more the modified information deviates from the model’s prior, the less likely the model is to prefer it."

6

1 Comment

Typically most #llm work we do is to reduce the responses down to JSON, but LLms even #llama3 do well with structured data as SQL. I’ve built entire schemas fairly rapidly in the early days of #genai

4

One of the most prevalent anti-patterns in tech is engineers saying, "I am going to make this easy for data scientists." Not only have I never seen this make anything easier, but no one ever consults data scientists about what they need. I get it. From an engineer's perspective, they are trained to abstract and make top-level APIs. But please let me train a model in PyTorch. Thanks.

6

1 Comment

To my AI/ML colleagues in Consumer Goods and Retail - please check out Brian’s profile

3

I just took a position in Nvidia. I missed out on this first runup and was waiting on an entry point to get into the AI trade on this stock. My Capital Asset Model, Price to Earnings Model and my Momentum Model are all forecasting a good financial return on this stock. However, my Free Cash Flow Model is estimating that this stock is worth $1,822.75. Which is more than double what the stock is trading at now. So this stock, as of right now, has the potential of a multi-year hold. I would estimate from my models that some exit points in this trade are: $940 <- Consensus Wall Street valuation, $1,419 < - My Momentum Model valuation and $1,889 <- My Price to Earnings Model valuation. There should be a reasonable expectation of earning a decent return on Nvidia inside of the next 12 months. By Amos B Robinson | www.RobinsonAnalytics.com

2

The entire essay is great. But the paragraph below really got me thinking. "First of all: what is work? Work is of two kinds: first, altering the position of matter at or near the earth's surface relatively to other such matter; second, telling other people to do so. The first kind is unpleasant and ill paid; the second is pleasant and highly paid. The second kind is capable of indefinite extension: there are not only those who give orders, but those who give advice as to what orders should be given. Usually two opposite kinds of advice are given simultaneously by two organized bodies of men; this is called politics. The skill required for this kind of work is not knowledge of the subjects as to which advice is given, but knowledge of the art of persuasive speaking and writing, i.e. of advertising." https://lnkd.in/g3iw9HDW

12

From Princeton & Microsoft There is a lot of interest in estimating LLMs' uncertainty, but should LLMs express uncertainty to end users? If so, when and how? v/ Sunnie S. Y. Kim In our #FAccT2024 paper, we explore how users perceive and act upon LLMs’ natural language uncertainty expressions. "I'm Not Sure, But...": Examining the Impact of Large Language Models' Uncertainty Expression on User Reliance and Trust https://lnkd.in/emqtjqPQ Through a pre-registered experiment (N=404), we examine how the presence and perspective (first-person/general) of LLMs' uncertainty expression impact user reliance, trust, and task performance, when users seek medical information with a fictional LLM-infused search engine. Through a pre-registered experiment (N=404), we examine how the presence and perspective (first-person/general) of LLMs' uncertainty expression impact user reliance, trust, and task performance, when users seek medical information with a fictional LLM-infused search engine. We also find perspective matters: LLMs’ uncertainty expressions in the first person show stronger effects than those in the general perspective. We view this as evidence that "language choices" matter in how users perceive and act on the outputs of LLMs. Our results suggest teams building and deploying LLMs and LLM-infused applications should evaluate approaches to mitigate overreliance, including language choices, carefully with users before release. We hope our work serves as an example and guide for such evaluations.

45

I came across two pieces on the current state and future of AI recently that would on the surface seem to be at odds, but are not so much when you consider the dynamics of the industry in a more nuanced way. The first is by acclaimed futurist Ray Kurzweil, who writes an optimistic piece in The Economist (pardon the paywall if you don't have a subscription) in which he doubles down on his 1999 prediction that we will have AGI by 2029: https://lnkd.in/gEwa6mfU As always, he offers thoughtful, informed analysis and no shortage of predictions about the future course of AI, some of which may actually prove accurate. The second piece, which has been making the rounds among the AI rank and file, is a Ludicity blog post which is as witty and funny as anything I've read in The Onion in its heyday: https://lnkd.in/gkmnf8DU Matt Sekerke, whose opinions on technology are consistently grounded and cut cleanly through the hype, pointed it out, and true to form, I found nestled among the humor and cynicism some real insights which are well worth considering by those who have otherwise been swept away by the AI hype cycle. Enjoy 🙂 #ai #artificialintelligence #agi #artificialgeneralintelligence #genai #generativeai

5

A lot of people think AI isn't explainable. It is.   If you build your model right, you can explain every forecast issued by AI -- for every location, every time horizon, every variable.   We use these explainability methods at Climavision to identify the patterns our models use to make their forecasts. We then use these explanations to help inform how we build our models and help customers trust our forecasts.   The plot below shows regions that are being used to make forecasts for the location shown by the blue dot. In this particular example, we are analyzing surface pressure patterns that are used to make a temperature forecast. The brighter regions are more important for making an accurate prediction on the date shown in the plot title.   Shoot me a DM if you're interested in learning more about our explainable AI forecasts. Our AI forecasts are specifically built for longer range forecasts from three weeks out to a year.

137

7 Comments

#MAJOR #AI 💡 break-through #SURFACED THIS WEEK 💬 While majority of the folks (and candidates I am having for job interviews) dream about riding the #LLM's horses of #GPT-type, in the background there are new trends really emerging. I have been watching closely work of Sepp Hochreiter and his team in giving #GenAI new hope by "resurrecting" LSTM. 🔭 For those of you, who don't follow the AI tech deeply, Long-Short-Term-Memory (or #LSTM) models have been the pre-cursors of the GPT technology. Most of the translators and other language bots being on market before launch of ChatGPT have been actually some form of LSTM application. When OpenAI broke through with Generative Pretrained Transformer (hence GPT) line of models, they looked like good-bye kiss to LSTM. GPT could work with so many more tokens, achieved better precision in tasks and was also faster in inference. 🪜 However, Sepp and his team recognized also one strong weakness of GPT family of models. Their learning is #quadratic, which makes push for more and more parameters prohibitive expensive. That also means that GPT's already got close to their (reasonable) ceiling of how big single model can be. (This is reason why they have been geared into round-of-experts model platforms, but that will be subject for another post of mine later). 💥Not succumbing to GPT dominance, Sepp Hochreiter worked hard on upgrading the LSTM by adding #exponential #gating, #matrix #memory and parallel training. This week they published paper on #xLSTM brand new generation of LSTMs that actually competes well with GPT-technology and are much cheaper to train. And that is after much of the hype a real technological advancement. Large Language Models are not peek of AI, they are just transition to another phase. And this whitepaper is major step on that pathway. Read full detail here :

5