Runtime’s Tom Krazit on the Beauty of Databases and Why He’s Not Concerned About Generative AI

Tom Krazit’s enterprise coverage is a fun read. He makes an inherently dry subject like cloud infrastructure interesting with his trademark dry humor and smart, biting analysis. I’ve been following his writing since we worked together at CNET more than a decade ago and he does a great job helping people make sense of enterprise technology and the companies that sell it. 

That’s why it was particularly disheartening when Protocol shut down last November, ending his more than three-year run there. With the spirit of an entrepreneur, Tom regrouped and in April launched his latest venture — Runtime. It does not disappoint. I reached out to Tom to see if I could get the skinny on Runtime and some of his thoughts about the state of enterprise tech. Here is an edited version of our Q&A.

Tell us about how you got into tech journalism. 

I started out covering tech at IDG, which played an enormous role in training a whole generation of journalists who have gone on to do amazing things, like Ashlee Vance and Bob McMillan. It was a fundamental place for me to learn what tech is – from the perspective of the people who have to pay for it and manage it. That has really centered a lot of my coverage. From that point forward, I’ve been trying to be the advocate for the person who has to deal with all this technology that vendors are cramming down their throats. 

So I started off there, and then there were twists and turns along the way. I like to joke that journalism is the world's slowest moving game of musical chairs. I've been pretty lucky to keep getting a chair during various downturns in the economy. Eventually I wound up at Protocol, which hired me in late 2019 to be their first cloud and enterprise reporter. Over the next couple of years, I built a newsletter and eventually had a team of six, until it all came crashing down last November.

Anyway, I wasn't sure what to do after that. I'd always thought about going out on my own and I was able to acquire the mailing list from Protocol Enterprise that I had built in my three years at the company. Now I’ve turned that into Runtime, which is a very similar type of publication. We're basically looking to do in-depth stories that a lot of other people in this space aren't doing — less product focused and more people focused, along with coverage of trends and emerging technology.

<split-lines>"[Runtime is] basically looking to do in-depth stories that a lot of other people in this space aren't doing — less product focused and more people focused."<split-lines>

What are you most interested in covering right now? 

Historically, I’ve covered infrastructure and the cloud providers, and traditional infrastructure approaches, like whether the on-premises renaissance is really a thing or not. I also don't think there's any way that we can ignore what's been happening with AI this year, and that obviously has filtered into everything that we cover. I'm perhaps a little bit more skeptical than some about where it actually winds up, but there's no doubt that it will end up in a lot of products that can influence this market. So, that's a huge story for sure. I'm actually working on a story about how Generative AI has created all kinds of new worries – for instance, corporate data getting leaked when large language models are hooked into sensitive data. 

How big of a story is Generative AI for developers?

I think a lot of issues have to be ironed out before enterprises feel super comfortable with a lot of these tools for software development, but they're coming. It’s like the old shadow IT issue. Your developers will be using it whether you authorized it or not. It’s a force multiplier. I was pretty skeptical about the whole thing until a couple of months ago when Steve Yegge (head of engineering for Sourcegraph) wrote a piece about how we don't trust our human developers anyway. We have to check their work. AI is going to screw up, too, but it's fine because we have a procedure for fixing these things. And the light bulb kind of went off for me. The speed at which Generative AI operates is so much faster and you have to do the debugging anyway, so what's the difference? I think ChatGPT falls down when it tries to pretend to be a human. And I think that we are not going to be replaced that quickly by these things.

<split-lines>"I think ChatGPT falls down when it tries to pretend to be a human. And I think that we are not going to be replaced that quickly by these things."<split-lines>

What about more traditional enterprise tech stories you are following?

The broader topic is data management and how big data has evolved into modern data warehousing. I think companies like Snowflake and Databricks have really taken this industry in a different direction that way. A 20-year-old me would punch me for saying this, but databases in general have been super interesting over the last couple of years because there's been so much yield. For a very long time, there were maybe three databases, and now there's like 40 and they are purpose built. That’s really interesting as a technology buyer. It's also really confusing for tech buyers because you have to sort through all these different things. There's a lot of territory here to explore. 

How do you want to be pitched?

If you don't hear from me, I'm probably not interested. I can’t respond to every email. If you pitch me a topic that we have talked about in the past or that you know I'm interested in, I don't mind a follow up in that case, because I do miss emails. I try to respond quickly when it's something that I’m interested in. What I have found pretty useful is to have emails that say, ‘Hey, I can put you in touch with these eight companies,’ so it's searchable in my inbox. And then when I have an idea for a story, I can find a source pretty easily.

<split-lines>"If someone started a bunch of companies, they're from an underrepresented background in tech, or they're an interesting technologist who made some breakthrough, I want to meet those people."<split-lines>

What compels you to open a pitch or to consider covering a story? 

I get asked this question all the time and I wish I had a better answer, but the only thing I can say is it just has to be good. It has to be a company or a person that's interesting or a technology that's compelling. If it's something I've been hearing people talk about who don't have a commercial interest in it, then that's going to make me sit up. For profiles, if someone started a bunch of companies, they're from an underrepresented background in tech, or they are an interesting technologist who made some breakthrough, I want to meet those people. But I'm probably not going to do the exact story around that person that you or your client wants me to do. I'd rather figure that out for myself.

More posts

July 24, 2024

July 22, 2024

Expert Insights
Life Sciences

'The Pipeline' Podcast - Episode 3: Breaking Through the Crowded AI Media Landscape

July 18, 2024

July 16, 2024

Inside Mission North
Growth

Must-Have Integrated Communications Goals for Every Stage of Growth

July 15, 2024

July 15, 2024

Expert Insights
Enterprise

Sentry’s Director of Communications on Marketing Authentically to Developers

July 19, 2024

July 11, 2024

Expert Insights

How Executive Thought Leadership Builds a Brand