(This article is an excerpt of Chapter 5 from the book "Tech Fluent CEO" — a new breakthrough book for non-technical people who want to build and lead digital companies.) In most system diagrams, a database is depicted as a cylinder (as shown in the figure) that just sits there as an opaque black box, simply storing data. (Looking forward to your fan letters for my award-winning drawing and calligraphy skills.) In reality – and you might laugh at me for saying this – modern database systems are so cool and sophisticated that I almost see them as living,…
(Disclaimer: this essay is an excerpt from my book, Tech Fluent CEO — for "non-technical" founders and professionals who hate calling themselves that. You can get it here for free by paying with a tweet, or purchase it here. Also available on Amazon.) There’s a huge communication gap between “technical” and “non-technical” people in the industry, even in Silicon Valley. Non-technical professionals are constantly told that the path to “technical fluency” is through learning to write code or some other kind of intense education. That’s terrible advice. It's good to give it a try and see if you enjoy it,…
This is about AlphaGo, Google DeepMind’s Go playing AI that shook the technology world in 2016 by defeating one of the best players in the world, Lee Sedol. Go is an ancient board game which has so many possible moves at each step that future positions are hard to predict — and therefore it requires strong intuition and abstract thinking to play. Because of this reason, it was believed that only humans could be good at playing Go. Most researchers thought that it would still take decades to build an AI which could think like that. In fact, I’m releasing this essay…
Google’s DeepMind is one of the world’s foremost AI research teams. They’re most famous for creating the AlphaGo player that beat South Korean Go champion Lee Sedol in 2016. The key technology used to create the Go playing AI was Deep Reinforcement Learning. Space Invaders! Let’s go back 4 years, to when DeepMind first built an AI which could play Atari games from the 70s. Games like Breakout, Pong and Space Invaders. It was this research that led to AlphaGo, and to DeepMind being acquired by Google. Today we’re going to take that original research paper and break it down…
I promise you won’t have to use either Google or a dictionary while reading this. In this post I will teach you the core concepts about everything from “deep learning” to “computer vision”. Using dead simple English. You probably already know what self driving cars are and that they are considered the dope shit these days, so if you don’t mind I’m going to skip any high school essay-ish introduction. 🙂 But I’m not skipping my own introduction: Hi I’m Aman, I’m an engineer, and I have a low tolerance for unnecessarily “sophisticated” talk. I write essays on Medium to…
Even Nobel Laureates agree that scientific papers are headed towards jargonocalypse. <Rant> I am an engineer by education, and have therefore studied science a lot. Yet I have realised that I find it easier to read something written in Spanish, than most scientific papers written in English (and I’ve never even fucking learned Spanish). Do you know why scientific research papers are so hard to read? I’m going to be unfairly brutal here. Because the authors of these papers these days, the “global scientific community”, predominantly consists of people whose life depends on one thing: whether they can coat their…
Hint: It’s not in America or Europe. Earlier this week I watched an amazing documentary by Wired about the rise of Asia’s silicon valley hotspot in Shenzhen in China. This one small city, which came into prominence just two decades ago and doesn’t have a lot of history or culture of its own, is now the breeding ground for technology innovation in the East. Shenzhen is the crystal ball in which you can gaze at the future, even more than Silicon Valley. And the most fascinating thing about it is that this city breeds hardware innovation, not just software innovation.…
(For Mac users) I’m an engineer. I hate most of the documentation I find on the internet. I guess developers write documentation for each other, which is a pretty understandable sentiment. And we use lingo that makes sense to our club. But unintentionally, often this super technical language keeps out beginners who are just taking an online course or reading a book and just want to get things done. I had a similar experience while trying to set up Google App Engine while taking Udacity’s Full-Stack Web Developer Nanodegree. Udacity makes it a point to not spoon-feed everything to students…
When I arrived in China, I realized I was on WeChat all the time. I was using it in many different ways: Instant messaging A social network like Facebook, for sharing pictures/videos/links/etc Sending money to people Making audio and video calls over the internet Quickly exchanging files between my phone and desktop (Wechat’s desktop and web clients automatically create a “file transfer” conversation, just like any other conversation with a contact. By sending a file to this conversation, you instantly see it on the other end) As my personal visiting card (to share my contact details with anyone, I simply…
This is my opinion of course, but I believe Reinforcement Learning is where the most formidable potential of AI lies. In this video, I give a simple explanation and food for thought.
Everyone knows about "company culture." This video is not about that. Engineering culture is specific to tech teams, and it is much more "concrete" and tangible — it can also be controlled to a greater degree. If you're a non-technical CEO, it would be very helpful for you to understand this concept. Let me give you an introduction!
"Safe AI" is a trending topic in the AI community recently. If you're a CEO, should you really care? My 2 cents: not any more than how much you already do. But you should still watch this video, to understand what "safe AI" means. Developing AI applications, in their current form, is not THAT different from any other software. It needs to be verified and validated according to the business objectives. The devil is in the details, as we discuss.
AI has an immense number of industrial applications. One of them is in noticing when things go wrong when humans don't, and also predicting WHEN things are likely to go wrong. The key technique here is called "anomaly detection." In this simple video, I explain what that is!
A sensitive topic, but somebody's gotta talk about it. Whenever you have a big technical communication gap within a company, it needs to be managed somehow. In this video, I share some ideas and concepts that will help you navigate the delicate conversations that happen.
I've got some crazy ideas here that I'd like you to consider. A lot of tech companies instinctively create a team of ML engineers, believing that they will do cool AI stuff that the rest of the company will benefit from. This is a big mistake if you actually want to build AI products. Here's why, and what you should do instead!
Data-Centric AI is the new trend in the machine learning community, meant to accelerate the adoption of machine learning in various industries — out of the research lab and into the real world. In this video, I explain what it is, in simple words!
Hiring is hard, especially if you're recruiting someone who is from outside your domain of expertise — let's say you're a non-technical founder hiring your first developer or CTO etc. Today's deceptively simple tip is that peak performers are obsessed about peak performance! When looking for a truly world-class professional, ask them about their ROUTINES!
Edge AI is quickly becoming a rage. In this video, I give a friendly introduction for non-technical people! For reference, here are two relevant podcast episodes we've published that deal with Edge AI: Using ML to build devices that get better on their own: Ekkono The next generation of specialized chips for AI: Kneron
