12 October, 2024

My POV on AI Trends vs Crypto Trends

A comparison of the recent AI and cryptocurrency trends

My POV on AI Trends vs Crypto Trends
Available in:
 English
 Vietnamese
Reading time: 5 min.
Table of content

    As someone who's been in the tech industry for a while, I've seen my fair share of trending cycles. Two of the most prominent in recent years have been cryptocurrency and artificial intelligence. Today, I want to share my perspective on these two phenomena and why I believe the AI trends is more substantive than the crypto boom we witnessed.

    dogecoin meme

    The Crypto Trends: The chaos

    Let's rewind to the height of the crypto craze. It was a time when the world was grappling with COVID-19, economies were in flux, and uncertainty was the only constant. In this chaotic environment, cryptocurrency seemed to offer a golden opportunity:

    • People with money were desperate to invest it somewhere.
    • Those without were looking for a quick way to make it.
    • The promise of decentralized finance was alluring in a world that felt increasingly unpredictable.

    But here's the thing: amid all the excitement about Bitcoin prices and NFTs, how many people were actually discussing the underlying blockchain technology? In my experience, not many. The focus was overwhelmingly on financial speculation rather than technological innovation.

    Image: Graph showing Bitcoin price volatility during and after the pandemic

    bitcoin price volatility

    The AI Revolution: Solving Real Problems

    Before I dive deeper into my thoughts on AI and crypto, I think it's worth sharing a bit about my own journey in the world of machine learning. It's a path that's shaped my perspective on the current AI boom.

    My Journey in Machine Learning

    My first encounter with machine learning was back in high school in Vietnam. I remember stumbling upon VGG16 (an image recognition model) and being instantly fascinated. But let me tell you, it wasn't easy to learn about ML back then. The field was pretty obscure, especially in Vietnam, and there weren't many resources available online.

    My big break came when I moved to France. While I was studying cybersecurity, I landed an apprenticeship with an AI-powered chatbot company. Talk about a crash course in applied machine learning!

    In my free time (what little I had), I dove into open-source projects. One of the most exciting was contributing to llama.cpp. If you're into ML, you probably know how big of a deal that project is.

    After graduating, things moved quickly. Within a year, I received an offer from Hugging Face. For those who might not know, Hugging Face where all the cool open-source AI stuff happens. Getting that offer felt like validation for all those years of self-study and late nights coding.

    And now? Well, I'm living the dream, honestly. I'm a full-time employee at Hugging Face, working on open-source AI and ML projects. And I'm still contributing to llama.cpp on the side. It's wild to think about how far I've come from that high school kid in Vietnam, struggling to find ML resources online.

    The current state of AI

    Now, let's contrast this with the current state of AI. The difference is stark:

    Direct Problem Solving

    LLMs are tackling real-world challenges head-on. Remember the days of painstakingly coding text patterns for simple Hello, how are you chatbot responses? Or maintaining separate models for each language pair in translation systems? Those days are behind us. One LLM can now handle multiple languages and complex queries with ease.

    For example, imagine you're trying to build a smart assistant for a local business. Before AI, here's what you'd have to do:

    • Write a ton of specific rules for every possible question a customer might ask.
    • Create separate systems for different languages.
    • Constantly update these rules as products or services change.

    Now, with Large Language Models (LLMs), it's an easy game - just look at Zendesk's ChatGPT integration example.

    Broad Impact

    It's not just about chatbots. AI is transforming fields we once thought were the realm of science fiction:

    • Robotics that can adapt to complex environments
    • Self-driving cars navigating busy streets
    • AI-generated art and videos that blur the line between human and machine creativity

    Image: Tesla "Optimus" Robot

    Tesla Optimus Robot

    Developer Productivity

    As a developer, I can't overstate how AI is revolutionizing our workflow. From generating test cases to assisting with refactoring, AI is taking care of the repetitive tasks, allowing us to focus on more creative and complex problems.

    Image: Github Copilot - The AI chatbot that can help developers to write code

    Github Copilot

    Ongoing Discovery

    Even after nearly two years since ChatGPT's release, we're still uncovering new capabilities and applications. For a long time, researchers all consider machine learning models as black boxes. Not so long ago, we started to see some cool experiments like Golden Gate Bridge Claude - it is a perfect example of how we're just beginning to understand and extract the full potential of these models.

    The Bottom Line

    While both AI and crypto have generated significant hype and trends, I believe AI's impact on solving real-world problems makes it more substantive and promising. It's not just changing how we invest or speculate; it's fundamentally altering how we work, create, and interact with technology.

    What's your take? Have you experienced the impact of AI or crypto in your work or daily life? I'd love to hear your thoughts in the comments below.

    References

    Want to receive latest articles from my blog?
    Follow on