Checksum Logo

Three Stages of Technology Transformation

Gal Vered
Explore the transformative power of LLMs in Checksum's insightful guide to the Three Stages of Technology Transformation, showcasing how AI revolutionizes software testing.

At Checksum, large language models (LLMs) are a core technology that we use to generate and maintain end-to-end tests for our clients.

LLMs are still exploding in popularity, and we consider them to be a breakthrough technology with vast potential beyond their current use.

This is the mental model we’re using internally to think about how LLMs will be used in the future – and how our product (and similar LLM-based products) will evolve. We call this the Three Stages of Technology Transformation.

We’re in the early stages of a technology transformation around LLMs, but we have seen this pattern unfold many times before. Below, we’ll analyze how the internet disrupted the news industry, and make some predictions about how our own product roadmap will unfold.

Stage 1: Cheaper and faster

In the first stage of technology transformation, innovators are just getting the hang of the new technology. The most common pattern is applying the new technology to existing needs in a way that is faster, cheaper, or both.

Consider how the early days of the consumer internet disrupted the news industry. News orgs like the Columbus Dispatch, the New York Times, the Washington Post simply put their news online – as well as individuals who reposted similar content onto blogs and internet-only news websites.

This was a faster and cheaper way to distribute and consume news, but the core news product was unchanged: the articles were just posted online.

At Checksum, our core test generation feature looks quite similar. Before Checksum, writing software tests was a time-consuming process that only engineers could do. With Checksum, it’s possible to generate tests more quickly and cheaply using AI.

Stage 2: New possibilities

In the second stage, innovators begin to discover totally new applications of the new technology. Totally new options become available – and use of the new technology looks less like a copy-and-paste, and looks more like a new paradigm.

For the internet and news industry, stage two started with Facebook’s launch of the News Feed: a real departure from the first stage of traditional-news-but-online. With the Facebook News Feed, for the first time, everyone got their “own newspaper” with endless scrolling, personalized to their interests.

For Checksum, we crossed into stage two with the launch of real-time autonomous test maintenance: a totally new application made possible by LLMs. Using AI, we’re able to ensure that a test suite is continuously accurate without flakiness.

As software changes, tests are bound to break and be wrong. And every time a tests breaks, an engineer needs to fix it – but it can only be fixed after the fact. With this new approach, we are able to fix tests during the run in real time, so the results are accurate from the start – something that was simply not possible before.

Stage 3: New playing field

The shift from stage two to stage three is harder to pinpoint, but at some point, the breakthrough technology is so meaningful that it completely redefines the original field.

Facebook and social media completely changed how people consume content – and likely much more. The smartphone completely changed how people interact with computer software.

For Checksum, stage three is still a vision, and we have a long road ahead. Given that the cost of generating a test is practically zero, and tests are always accurate, we can transform the way that engineers write and ship code.

If you can generate unlimited tests and run them accurately every time, we can ensure quality at every step of the way. Checksum will get a detailed analysis of how bugs are created, manifested, and fixed.

As an engineer writes code, Checksum will pull the most critical tests from an unlimited reservoir and provide feedback in real time. Bugs will be detected as they are written and fixed on the fly. When bugs are fixed, they won’t be fixed in a superficial way – deeper architectural flaws and underlying tech debt will be addressed, too.

It’s an ambitious vision for sure, but hey, autonomous agents that can learn from users and test your app’s E2E tests was also just a vision one year ago!

About The Author

Gal Vered is a Co-Founder at Checksum where they use AI to generate end-to-end Cypress and Playwright tests, so that dev teams know that their product is thoroughly tested and shipped bug free, without the need to manually write or maintain tests.

In his role, Gal helped many teams build their testing infrastructure, solve typical (and not so typical) testing challenges and deploy AI to move fast and ship high quality software.