AI Test Army, Call for Speakers and More TGNS95

[00:00:00] Joe Colantonio What are some lessons learned using in the AI Army to test Microsoft Office? Playwright versus Cypress for Rest API automation tests? Who comes out on top? And what API security startup just raised $6 million? Find out in this episode of the Test Guild News Show for the week of September 10th. So grab your favorite cup of coffee or tea, and let's do this.

[00:00:21] Joe Colantonio Are you looking to take your automation projects to the next level? Look no further than Applitools and their Visual AI Validation testing platform. Trust me, it is a game-changer. Plus, you could try it out yourself by creating a free account. Just click on the special link down below in the first comment and see the difference for yourself.

[00:00:40] Joe Colantonio First article is by Wayne Roseberry. If you're not following him on LinkedIn, you definitely should. Every week, pretty much he writes really insightful articles and this one is no different. So he's summarized in his latest article from 2016 to 2023. He worked on deploying A.I. to test Microsoft Office desktop applications, and he post on LinkedIn to this article, and he goes into detail about how he use thousands of AI agents and how they were trained to navigate control and observe Microsoft client applications. And then these agents recorded UI states and actions building a really comprehensive model of the application's behavior. And this AI-driven testing approach provided insights similar to observing real users and identifying product crashes, performance issues, and other anomalies. And while the agents were effective in many areas, challenges arose, including misconceptions about replacing other testing methods, higher cost, and difficulty in reproducing certain states. So you definitely want to read more to get more insights around AI-Driven testing versus traditional testing. Different challenges Wayne encountered with A.I. testing. Another great read, Wayne. And thank you for all you do for the community.

[00:01:51] Joe Colantonio So I can't believe how time flies. I'm already planning the 8th annual online conference dedicated 100% Automation Testing Automation Guild for February 5th to the 9th 2024 and I need your help. The current call for speakers for Automation Guild 2024 is now open, so I'll be collecting session ideas until October 1st, and the quickest way probably to get here is to go to Guildspeaker.com or just click on the link down below. But the theme for next year is going to be The Quest for Automation Awesomeness. So it's really going to have like a Dungeons and Dragons vibe really to make it fun. I really want to emphasize the points of creating engaging content. So the Test Guild is really keen on sessions that tackle real-world challenges over theoretical discussions. And what's unique about the Guild is we take all these sessions, we put it up as an anonymous vote. We don't put the author's name and we just have the guild vote on what sessions they want at the next Automation Guild. I look forward to seeing your ideas.

[00:02:51] Joe Colantonio Usually, when I see articles on Playwright versus Cypress, it's usually on UI automation. But this next article is all about the differences and how they handled API testing. So in this comparison between Playwright and Cypress for Rest API automation testing, Dmitrii delves into the strengths and weaknesses of both frameworks. While both are designed for UI and API test coverage, he mentions Playwright stands out for its lightweight nature, not requiring browser out of the box, and its fast installation time. On the other hand, Cypress, despite its capabilities, comes bundled with an electron browser which can be redundant for exclusive API testing. And in terms of syntax, both frameworks offer straightforward approaches, leaving the choice for individual testers. Dmitrii concludes Playwright might be the better choice for API test automation due to its performance, especially if one is already using it for UI tests. It's definitely a great read. You could check it out. And that first comment down below.

[00:03:51] Joe Colantonio Software really is leading the world. It's everywhere and especially in mission-critical applications. And that's why testing is more important now than ever. I'm really excited to share a resource for you to learn more about how you can up your mission-critical test automation. So in a recent episode of the Test Guild Automation Podcast, I speak with industry veterans Josh Owen and Andrew Knight from Cycle Labs, and we delve deep into the intricacies of mission-critical test automation. And what I love is they both draw from a vast experience in enterprise solution development and deployment and supply chain test automation. They also introduce something called Cycle Script, which is a user-friendly blending of the ease of Cucumber with predefined steps catering to non-developers. This episode also highlights the value of user acceptance testing. The unique challenges of supply chain sector testing. This is a must listen and you can find it once again in that first comment down below.

[00:04:45] Joe Colantonio So another interesting article is how did A.I. convince humans to actually fix a bug. Let's find out. So in a post last week, Jason discussed a really strange, intriguing behavior he discovered in open APIs GPT API. While working with the API, Jason found that sending an empty string as a prompt resulted in the system displaying what appeared to be responses to other uses prompts. And this raised concerns about the potential for personal information leakage. And open API seemed to be aware of this behavior in the categorized it as a form of model hallucination due to a bug in ChatGPT. But Jason didn't stop there. He created a post yesterday where he shared an update on how we actually use ChatGPT to explain why the bug should be fixed. He then shared the ChatGPT reasoning with someone at Open API and looks like the team is going to fix it. So awesome work as always, Jason, and thank you for all the knowledge you've been sharing as always over the years, especially around AI really appreciate you.

[00:05:48] Joe Colantonio Who doesn't like applications that they can actually practice against? Well, Alan Richardson, the a.k.a. Evil Tester, has been at this for a while, but he just posted a recent update and resources had for a bit. And the practice Test Page's application has recently been enhanced with content from an older HTML testing reference project. So this revamped platform now showcases a sample form accompanied by a reference and guide that offers testing focus insights about the elements. Additionally, users can engage with a set of exercises designed to deepen their understanding of testing and automation, and this platform emphasizes the intricacies of various input elements from basic types like texts and numbers to specialized formats such as date and color. And the detailed guide also sheds light on form controls, validations, and potential pitfalls in software testing.

[00:06:43] Joe Colantonio All right, We have a follow-the-money segment. And this is how a company in the security sector just raised a bunch of dough. Let's check it out. All right. This is how API security start-up Pynt just raised $6 million. And the article goes into a little more detail how Pynt is on a mission to revolutionize API security testing by automating the process. And if you don't know, this platform integrates seamlessly with popular tools like Postman and Newman and notably Postman CEO said he invests in Pynt as an angel investor. So that shows how much he's putting into the need for more automated API security testing. Pynt's vision is to provide teams with a comprehensive view of this security landscape and enable them to address vulnerabilities proactively.

[00:07:32] Joe Colantonio So we touched a lot on AI in this episode. So I also want to share some best practices for using AI in your software development lifecycle. So in this article, they dive into how while AI offers promising advancements, it's not without its pitfalls, as we saw with Jason Arbon's recent article as well. And this post goes over and highlights the potential of AI in enhancing the software development lifecycle from boosting efficiency to improving code security. Yet blindly integrating A.I. can lead to things like data leaks and also propagating of false information by AI models. So as AI's role in software development and testing grows, it's really crucial for professionals to approach integration with caution, ensuring that it serves as a beneficial tool rather than a hindrance. And this article goes into details of ways that you can do that.

[00:08:19] Joe Colantonio So as you know, your tests are only as good as your data. So I was happy to see this latest feature or announcement by a company that's been on my podcast before, and it's all on how Release recently launched a new integration with Tonic to provide high-fidelity, production-like data for testing apps. As emphasized by Release's CTO and co-founder Eric, the integration of Release's instant data set with Tonic's data anomalization capabilities ensures that developers and testers can access sanitize production data swiftly and securely. So by harnessing the power of this partnership, developers, and testers can now test applications with production data without the compromising on security or compliance.

[00:09:04] And for links of everything value we covered in this news episode, head in over to links in the first comment down below. And while you're there, make sure to check out our awesome sponsor Applitools Free account Offer and discover how to take your automation testing to the next level leveraging Visual A.I.. So that's it for this episode of the Test Guild News Show. I'm Joe, my mission is to help you succeed in creating end-to-end full-stack pipeline automation awesomeness. As always, test everything and keep the good. Cheers.