2 ms·
Three suggestions: 1. Embedded. The world is creating tons of new products with increasing levels of technological sophistication. Much of the libraries, tooli
by tlack 10y ago
Three suggestions:
1. Embedded. The world is creating tons of new products with increasing levels of technological sophistication. Much of the libraries, tooling, and learning materials needed are yet to be created. Start with Arduino on the ESP32.
2. IoT security. Devices like the ESP8266 and ESP32 are quickly finding their way into many different applications, but as far as I know, no one has vetted their RTOS. This is dangerous because they are networked and difficult for consumers to understand in a tangible way. Likewise how can physical security be protected with these chips and others like it?
3. Machine learning. Be an early adopter and discover ways to use this new technology. There's a great-looking JS ML framework that you can use to get your feet wet without jumping to R, C++, or Python. https://github.com/cazala/synaptic https://github.com/cazala/synaptic
- csthrw 10y agoThanks for your response. I've though about embedded and I've already bought an ESP8266 and played a few days with it maybe I should invest more time. I've already played a little bit around with machine learning, completed Andrew Ng's ML course. I'm actually really thinking about switching to ML and try to get a new job in a few months. But I don't want to just follow a trend because it's upvoted on HN. And I didn't really come across entry level ML jobs, all jobs require masters or phd.