Interesting Content in AI, Software, Business, and Tech- 05/15/2024[Updates]

Content to help you keep up with Machine Learning, Deep Learning, Data Science, Software Engineering, Finance, Business, and more

May 15, 2024

Hey, it’s Devansh 👋👋

In issues of Updates, I will share interesting content I came across. While the focus will be on AI and Tech, the ideas might range from business, philosophy, ethics, and much more. The goal is to share interesting content with y’all so that you can get a peek behind the scenes into my research process.

I put a lot of effort into creating work that is informative, useful, and independent from undue influence. If you’d like to support my writing, please consider becoming a paid subscriber to this newsletter. Doing so helps me put more effort into writing/research, reach more people, and supports my crippling chocolate milk addiction. Help me democratize the most important ideas in AI Research and Engineering to over 100K readers weekly.

Help me buy chocolate milk

PS- We follow a “pay what you can” model, which allows you to support within your means. Check out this post for more details and to find a plan that works for you.

I'll be in SF 24-27th and San Diego 28th to 30th. Let me know if any of you are around to meet then. We can dress up as insects, dance around bonfires, or do whatever else it is that the people in California do.

A lot of people reach out to me for reading recommendations. I figured I’d start sharing whatever AI Papers/Publications, interesting books, videos, etc I came across each week. Some will be technical, others not really. I will add whatever content I found really informative (and I remembered throughout the week). These won’t always be the most recent publications- just the ones I’m paying attention to this week. Without further ado, here are interesting readings/viewings for 05/15/2024. If you missed last week’s readings, you can find it here.

Reminder- We started an AI Made Simple Subreddit. Come join us over here- https://www.reddit.com/r/AIMadeSimple/. If you’d like to stay on top of community events and updates, join the discord for our cult here: https://discord.com/invite/EgrVtXSjYf.

Community Spotlight: Kiki's Bytes

Kiki's Bytes is a super fun YouTube channel that covers various System Design case studies. As an AI Engineer, I learn a lot from it since it teaches me about the important tools in software engineering at scale. This is crucial to me (and for many of you) who mostly interact with research and thus might struggle to deploy solutions. Absolutely elite resource for anyone technical.

If you’re doing interesting work and would like to be featured in the spotlight section, just drop your introduction in the comments/by reaching out to me. There are no rules- you could talk about a paper you’ve written, an interesting project you’ve worked on, some personal challenge you’re working on, ask me to promote your company/product, or anything else you consider important. The goal is to get to know you better, and possibly connect you with interesting people in our chocolate milk cult. No costs/obligations are attached.

Previews

Curious about what articles I’m working on? Here are the previews for the next planned articles-

Tech Made Simple

How To Make Less Dumb Mistakes When Programming (inspired by this video on adding rigor to programming).

AI Made Simple

KANs. A lot of you have asked for them, and I found their ideas pretty interesting.

Highly Recommended

These are pieces that I feel are particularly well done. If you don’t have much time, make sure you at least catch these works.

Another glorious battle for AI dominance… GPT-4o vs Google I/O

This is a pretty good summary of the announcements from Google vs OAI. I disagree that Google is behind. Seeing how this is the narrative adopted by notable commentators like Fireship, this is a clear indication of how bad their PR has been. I hope they fix it.

As a tangent, here are my 2 cents on the whole announcements: they were great and technically impressive announcements, but they weren't new. These capabilities have existed for a while. All the announcements will no doubt enable great products, but since we haven't heard about new capabilities- I'm guessing these teams are running out of steam. I hope we see this, and start investing some of the money being thrown at LLMs into other fields.

🔖 Topics Covered - Recap of Google I/O 2024 - GPT-4o features - Gemini 1.5 Pro features - What is project Astra? - What are the best AI models in 2024? - Latest features in Firebase - Firebase data connect - Firebase GenKit

Interpretable predictions of chaotic dynamical systems using dynamical system deep learning

We touched upon this in our piece on Chaotic Systems and AI, but I wanted to spotlight this specifically b/c it's got a very interesting approach and some good results.

Making accurate predictions of chaotic dynamical systems is an essential but challenging task with many practical applications in various disciplines. However, the current dynamical methods can only provide short-term precise predictions, while prevailing deep learning techniques with better performances always suffer from model complexity and interpretability. Here, we propose a new dynamic-based deep learning method, namely the dynamical system deep learning (DSDL), to achieve interpretable long-term precise predictions by the combination of nonlinear dynamics theory and deep learning methods. As validated by four chaotic dynamical systems with different complexities, the DSDL framework significantly outperforms other dynamical and deep learning methods. Furthermore, the DSDL also reduces the model complexity and realizes the model transparency to make it more interpretable. We firmly believe that the DSDL framework is a promising and effective method for comprehending and predicting chaotic dynamical systems.

Why You Hate The Modern World | Kierkegaard's The Present Age

I know I recommend Kierkegaard a lot. No, I will not stop. Unsolicited Advice is one of the best YouTube channels for explaining concepts in Philosophy/Literature. I attribute a lot of my critical thinking and writing skills to sources like this, and would strongly suggest you check them.

Kierkegaard's "The Present Age" or "On The Present Age" is his attempt to make sense of the changing world of 19th century Denmark, and the dawn of what we would now call modernity. And In this brilliant piece Kierkegaard takes us through so many issues we now view as major and established - predicting them over 150 years ago. So here is his guide to why you hate the modern world.

The Mushroom Motherboard: The Crazy Fungal Computers that Might Change Everything

This is so cool. I personally think that we're leaving a lot of performance on the table by not trying to integrate biological intelligence directly into artificial contexts (I remember seeing something a while back where a small piece of brain learned pong and other games a lot more efficiently than large RL Networks).

Unlock the secrets of fungal computing! Discover the mind-boggling potential of fungi as living computers. From the wood-wide web to the Unconventional Computing Lab, witness the evolution of mushroom technology.

How Venture Capital Make Decisions (TOP 20 Reasons)

A great overview by Rubén Domínguez Ibar . If you're looking to build your own startup, he's a great person to follow.

Shell's Fake Carbon Credit Scandal Explained!

This is another example of companies greenwashing and only paying lip-service to green efforts instead of making actual changes. We did a deep dive into that here.

Europe’s largest oil and gas company Shell was accused in an investigative report from Greenpeace Canada of selling millions of carbon credits tied to CO2 removal that never took place. Let’s look at what Shell did, how carbon offsets work, and how environmentally beneficial they actually are.

Recap of Text to Audio

A great survey by Roger Lam . I hope he keeps posting because me makes my life a lot easier.

Llama-3 agents that can browse the web by following instructions and talking to you

This is bittersweet for me. On one hand, I am very excited about this. On the other, it reminds me that Meta never released Sphere to the public. #ReleaseSphereNow

Our first agent is a finetuned Meta-Llama-3-8B-Instruct model, which was recently released by Meta GenAI team. We have finetuned this model on the WebLINX dataset, which contains over 100K instances of web navigation and dialogue, each collected and verified by expert annotators. We use a 24K curated subset for training the data.

It surpasses GPT-4V (zero-shot *) by over 18% on the WebLINX benchmark, achieving an overall score of 28.8% on the out-of-domain test splits (compared to 10.5% for GPT-4V). It chooses more useful links (34.1% vs 18.9% seg-F1), clicks on more relevant elements (27.1% vs 13.6% IoU) and formulates more aligned responses (37.5% vs 3.1% chr-F1).

📝 Guest Post: The EU AI Act – A Guide for Developers*

The Sequence puts out some bangers. I'm surprised this didn't get much attention b/c it was my fav of their work (and helps people understand important regulations).

The regulation is primarily based on how risky your use case is rather than what technology you use . The act splits AI applications into 4 possible risk categories: prohibited, high, low and minimal:

Prohibited AI Applications include systems that manipulate behaviour through subliminal techniques, exploit vulnerabilities, or engage in social scoring and real-time biometric identification in public spaces for law enforcement without strict justifications. Only governments can develop a prohibited application.
High-Risk AI Systems: These are involved in critical sectors like healthcare, education, and employment, where there's a significant impact on people's safety or fundamental rights. This is the category of application most affected by the regulation.
Limited Risk AI Systems: These are AI applications like chatbots or content generation. The main requirement here for compliance is transparency. The end-user should know that they’re interacting with AI.
Minimal Risk AI Applications: This category covers the majority of AI applications such as AI in games, spam filters and recommendation engines.

Artificial Intelligence Made Simple

Discussion about this post