Machine learning and AI have been revolutionizing the tech industry over the last several years. And with the emergence of new breakthroughs like generative AI, boundaries are continuously pushed and innovations are relentlessly driven.
On today’s episode of Startups On Demand, I am joined by Victor Dey, Tech Editor at VentureBeat. Victor combines his passion for writing and expertise as a data scientist towards becoming one of the most influential tech reporters in the media.
Today, we talk about all things ML and AI, the fascinating breakthrough of OpenAI’s ChatGPT, and the future of tech reporting.
Omri: Let’s start with your background. Can you talk about how you got into tech journalism and how you started your journey?
Victor: I always had a passion for writing. While I was doing my Bachelor’s Degree in Information Technology, I used to write blogs on the side, so that’s how it started. After completing my degree, I got an opportunity to work as a data analyst, and it was at that time that I understood the basics of data science, and how the technologies, architectures, and data work essentially. That’s how I found my interest and decided to pursue my Masters in Data Science.
Later on, I got another opportunity from a firm that reached out to me because I was one of the top students in class. They said, “Hey, we’re working on research based on data science, and I think you would be a good fit for us.” So I said yes and started working with them. And as you know, research requires writing your observations and heavy documentation, so I got a lot of experience while studying and understanding everything in detail. And that’s when I saw how interesting AI was, and I don’t think a lot of people know how these architectures work and how interesting they are.
After my master’s and when I started my first data science job, my boss reached out to me and said that they were looking forward to creating data science content, and that’s when they asked me if I would do tutorials. So I started writing a blog, and at the end of the blog, it had a quote that explained how this architecture works, which was a basic part of the tutorial. And after publishing my third or fourth blog, I started receiving so much love on LinkedIn that people started reaching out to me – students, teachers, CEOs, and professors from the biggest institutions. That’s what drove me further.
Omri: What is it about machine learning and artificial intelligence that you find super interesting?
Victor: I think the trends, models, and cognitive thinking that are being developed these days. We used to see things like these in the movies, but who knew that one day we will see self-driving cars? So artificial intelligence has gone to that level. Especially now, there’s so much power in the computing perspective. AI has become so strong that it can understand every track of humans. I think 10 or 15 years from now, we might really see how they will become really close to human thinking.
I have worked in NLP research, that’s why I find generative AI super interesting too. Deep learning and reinforcement learning as well. It’s just astonishing how robots, which are just merely algorithms, can remember what they did previously and they can tell you on that basis.
Omri: What are your thoughts on ChatGPT and how it can help writers?
Victor: OpenAI has been doing great, great work in the NLP models. We’ve already seen a lot of amazing legacy models from them like GPT, GPT1, 2, 3,, etc. Now, we’re seeing ChatGPT. I think they have a really good background in developing models like these, and I have been personally experimenting with ChatGPT for a few things, and it just gives you astonishing results. It’s unbelievable that the training of that model is so strong and it can literally answer you about your life. But right now, I don’t think any model can replace human intelligence and thinking, because the two of the most essential things that humans have are a) critical thinking/decision making and b) emotions. AI still lacks those. To get close to that, we would still need 10 to 15 years more. I don’t think it will replace writers because of our understanding, perspective, critical thinking, and opinions – there’s no match.
Omri: Other than generative AI, what other tech trends are you reporting or investigating in 2023?
Victor: So right now, what I’m seeing is that AR (augmented reality) and VR (virtual reality) will be heavily invested. We are also seeing process automation. I look forward to those things. Other than that, architectures are getting stronger. We have AI as a service, and network as a service. Right now, I also think metaverse fashion is developing, NFTs are coming in, blockchain will be integrated into metaverse soon, and that transaction engine will be stronger. I think it will streamline a lot of processes because what I’m seeing right now with companies like Walmart is that they’re training their employees with virtual reality and metaverse. I was also reading about how fire departments are training their workers with VR by creating real-life scenarios. So we can see heavy investments in that.
Omri: Are there any tech personalities that you’ve been thinking about recently, or that you want to interview?
Victor: The people who I most looking forward to interviewing are Yann LeCun who is into meta, Sundar Pichai, and Elon Musk definitely. I think they have a lot of interesting perspectives on a lot of things and I think these personalities work on the ground level. Although I’ve been very lucky to interview a lot of people and host a lot of webinars, these are the three people I most look up to as a data scientist.
Omri: How do you think the role of tech reporters has evolved throughout the years?
Victor: I think right now, people are open to new perspectives, to read, and understand your opinion. I think that gives us a lot of freedom to write, and people are embracing that because they just want to have a different perspective on things. In the past, people did object to a few things, but right now, people are ready to understand the criticisms. That is something that I’m absolutely using right now. Also, the love for AI – even from non-technical people. I look forward to helping even layman people understand better how this tech works and what it is so that when they finish the article, they could say “ah, we have understood something.”