For the past month, I’ve been tinkering with different new technologies, which have essentially made my life easier and 50% more productive. This is so far the best tech stack I’ve ever had, and I really feel I’m making the most out of it.
I’m not sure I’ll stick with MacOS for the future, and I still have a lot of concerns and issues about privacy, specially because of my recent extensive use of AI.
About the Threat Model
I really think this is an epoch in which the balance between usability and security/privacy is more important than ever, and can be very polarizing, specially with the use of AI.
From suggesting the next line of code in a terminal to chatting with us about quantum physics, it’s like having a little piece of the future in our hands. But, as with all great things, there’s a “but.” Before we dive headfirst into using AI everywhere, it’s super important to pause and think about our security. This is where threat modeling comes into play.
It’s a fancy term for basically asking, "What could go wrong?" and "How can we prevent it?" By understanding the risks specific to how we’re planning to use AI, we can make smarter decisions that keep us safe while we explore all this cool new tech. So, it’s not about being scared of using AI, but more about using it smartly and safely. Because, let’s face it, who wants to deal with a digital mess when we could be having fun exploring all the amazing things AI can do? And that means, personal responsibility (A term I’ve been having in the back of my mind about my own usage of services and whenever I recommend anything to anyone).
So, with that in mind, let’s get list the awesome tools I’ll be using for the rest of this month:
Terminal, CLI Tools:
- Termbot AI - My own CLI AI tool with A LOT of functionalities.
- Daniel Miessler’s Fabric - Incredibly useful and smart tool for CLI interaction with AI.
- Codeium for my IDE (It’s really amazing, although there’s no Chat capability yet for Vim/Nvim). It’s not great for Markdown files but it’s not meant to be used as an LLM but as a code completion engine.
- Warp Terminal - I just replaced Terminator with Warp, and it took me a while to get accustomed to it, but now it really is my go-to for all my terminal needs. I’m still waiting for Windows/Linux support. (Yes, it has AI)
Browsing/Internet:
- Arc Browser - Life changer Browser. (Yes, it has AI)
- perplexity AI - Incredibly useful and awesome AI-powered chatbot with internet search capabilities. Its API is (at least until now) very limited, and does not “chain” online lookups like the web app does.
LLMs | Large Language Models
- OpenAI API - I have yet to figure out how to get past/workaround Token limits.
- ChatGPT - Everyday copilot.
- Anthropic’s Claude 3 Opus specially for its API. I haven’t really done anything crazy with it yet, but I’m really excited about really putting it to use. It’s a little bit more expensive than GPT4 API.
Home Lab / Networks
- Tailscale - It REALLY fits all my needs for my basic homelab experience (No more bulky VPNs < Overlay Networks). The latency is very low, and for general use it’s free.
GPTs I’ve built this month:
- Risk Analysis GPT - Generates a “Top 5” risk approach for a given detailed prompt.
- Threat Modeling GPT - Generates a Threat Model from a given detailed prompt.
- Prompt Engineer GPT - Improves our prompt (Very useful).