Will AI Take QA Jobs | Best Test Automation Tool To Learn - Part 1

Summary

In this video, Alex interviews Aram B, a QA instructor, about the impact of AI on QA engineers' jobs. Aram discusses the rise of AI-based test automation tools like Test Trigger and how they don't threaten QA jobs but empower engineers. He advises engineers to embrace AI, learn to use it, and not to panic about job loss. Aram emphasizes the importance of understanding AI's limitations and the necessity of human oversight to prevent issues like the CrowdStrike outage.

Takeaways

• 👨‍🏫 Aram B is a QA instructor who teaches courses on becoming a software developer using Playwright.
• 💻 Aram's course is praised for its high quality, easy-to-follow content, and comprehensive slides.
• 🌐 The course is available on his website, bonderacademy.com, and is affordable.
• 💼 There's a concern that AI might take over jobs, particularly in Quality Assurance.
• 🤖 Aram suggests that QA engineers should not fear AI but rather learn to use it to their advantage.
• 🔧 AI tools like Test Trigger are emerging, allowing for codeless test automation through prompts.
• 📈 The trend shows an increase in AI-based tools for test automation, with more companies developing such technologies.
• 🛠️ Despite the rise of AI, the importance of understanding coding basics remains crucial for engineers.
• 👀 Aram emphasizes the need for human oversight in AI automation to catch errors and false positives.
• 🔎 AI can speed up coding with tools like Copilot, but it may also introduce errors if not properly managed.
• ⏱️ Elon Musk's approach to AI training involves using data before AI's creation to ensure quality and relevance.

Q & A

Who is Aram B and what is his profession?

Aram B is a QA instructor who specializes in software testing and automation.

What course did Alex take from Aram B?

Alex took a course on becoming a software developer using Playwright, a modern tool for automation.

What is the name of Aram B's website?

Aram B's website is called 'bonder academy.com'.

What is the price of the course mentioned in the transcript?

The course can be purchased for 100 dollars, which allows access to the whole course content.

What is the concern that QA engineers have regarding AI?

QA engineers are concerned that AI might take over their jobs, specifically in the field of quality assurance.

What is Aram B's opinion on AI taking over QA jobs?

Aram B believes that engineers should not fear AI taking their jobs, but rather fear those who use AI effectively. He suggests that AI is just a tool and it's the people who leverage AI that gain a competitive advantage.

What is 'test trigger' and how does it relate to AI?

Test trigger is an AI-based platform for test automation that allows users to automate scenarios without writing a single line of code.

What advice does Aram B give to those who are worried about AI replacing their jobs?

Aram B advises people to learn how to use AI and leverage its capabilities to boost their effectiveness and productivity.

What is Aram B's perspective on codeless automation tools?

Aram B thinks that while codeless automation tools can be a good starting point for understanding automation, professionals will likely still need to code to write complex automation scripts.

What is the importance of understanding the basics of coding according to Aram B?

Aram B emphasizes that understanding the basics of coding is crucial for connecting different pieces of code to work together effectively and for troubleshooting issues.

What is Aram B's view on AI-generated content and its potential issues?

Aram B points out that AI-generated content can sometimes be inaccurate or 'hallucinate' due to the data it was trained on, which may not always be reliable.

What example does Aram B give about the potential risks of relying solely on AI in testing?

Aram B cites the example of the CrowdStrike outage, where an AI-driven test automation tool reported a false positive, leading to a system failure that went unnoticed.

Outlines

00:00

📚 Introduction to Playwright Course

In this segment, Alex introduces Aram B, a QA instructor, and discusses the quality of Aram's course on becoming adept with Playwright, a modern tool for automation. The course is noted for its high-quality recording, easy-to-understand explanations, and comprehensive slides. Aram's webpage, bonderacademy.com, is highlighted as a resource for those interested in Playwright. The course offers an affordable plan for $100, allowing unlimited access to the entire curriculum. The conversation then shifts to a discussion about the impact of AI on the job security of QA engineers, with a focus on the potential of AI tools to automate testing processes.

05:00

🤖 AI in Test Automation

Aram shares his insights on the role of AI in test automation, mentioning his recent course on mastering Test Trigger, an AI-based platform that enables users to automate test scenarios with simple prompts, eliminating the need for coding. He discusses the increasing trend of companies developing AI-driven tools for automation and the importance for engineers to embrace AI to stay competitive. Aram advises against fearing AI but instead to learn how to use it effectively, drawing a parallel to the early days of Google search and the importance of understanding how to use AI tools to enhance one's skills.

10:00

🛠️ AI's Limitations in Coding and Automation

The conversation continues with a discussion on the limitations of AI in coding and automation. Aram explains that while AI can generate code, it often lacks the flexibility and understanding of complex logic required for professional-level automation. He emphasizes the need for engineers to have a strong foundation in coding to effectively manage and connect the pieces of code generated by AI. Aram also touches on the issue of AI 'hallucinations,' where AI provides incorrect information based on flawed assumptions or training data. He stresses the importance of human oversight to catch these errors and ensure the accuracy of AI-generated outputs.

Keywords

💡QA Instructor

A QA Instructor is an expert who teaches Quality Assurance (QA) practices and methodologies. In the video, Aram B is identified as a QA instructor who teaches courses on becoming adept with Playwright, a modern tool for automation. This role is crucial as it helps software testers and developers enhance their skills in ensuring the quality and reliability of software applications.

💡Playwright

Playwright is an open-source Node.js library created by Microsoft for automating Chromium, Firefox, and WebKit browsers with a single API. It is highlighted in the script as one of the most modern tools for automation. Aram B offers a course on becoming skilled with Playwright, emphasizing its importance in the QA field.

💡Software Developer

A Software Developer is a professional who designs, codes, tests, and maintains software applications. The video discusses how software developers, particularly those interested in QA, can benefit from learning Playwright and other modern automation tools to improve their efficiency and effectiveness in the software development process.

💡Automation

Automation refers to the use of technology to perform tasks with minimal human intervention. The video script talks about the increasing role of automation in QA processes, with tools like Playwright and AI-based platforms like Test Trigger, which allow for the automation of test scenarios without writing code.

💡AI

AI, or Artificial Intelligence, is the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. The video discusses concerns about AI potentially taking over jobs, specifically in QA engineering, and explores how AI tools are being developed to assist in test automation.

💡Test Trigger

Test Trigger is described as an AI-based platform for test automation in the script. It is an example of how AI can be leveraged to automate testing scenarios by simply writing prompts, eliminating the need for coding. This showcases the potential of AI to revolutionize the QA industry.

💡Coding

Coding is the process of writing computer programs in a specific programming language. The script mentions that while AI can assist in writing code, a deep understanding of coding is necessary for engineers to connect the pieces together and ensure the software functions correctly.

💡False Positive

A False Positive in the context of testing refers to a result that indicates a failure or error when there is none. The video discusses an incident with CrowdStrike's cybersecurity software, where an automation system reported a false positive, leading to a system outage. This highlights the importance of human oversight in automated systems.

💡Self-Healing Test

Self-Healing Test is a concept mentioned in the script where tests can supposedly fix themselves when they fail. This is an emerging feature in AI-powered automation tools, but the script raises questions about how such a feature would distinguish between a true bug and a situation that requires healing.

💡AI Hallucinations

AI Hallucinations refer to the phenomenon where AI provides incorrect or nonsensical outputs based on flawed or misunderstood inputs. In the script, this is discussed in the context of AI tools that may generate code or data that doesn't align with the actual requirements of a project, necessitating human intervention to correct these errors.

Highlights

Introduction to Aram B, a QA instructor who teaches a course on becoming a software developer using Playwright.

Praise for Aram B's course quality, with easy-to-follow content and slides provided.

Recommendation to check out Aram B's webpage, bonder academy.com, for affordable course plans.

Discussion on the concern that AI might take jobs, specifically in Quality Assurance.

Mention of AI tools for automation and how they are becoming more prevalent in the industry.

Aram B's view that QA engineers should not fear AI but rather learn to use it to their advantage.

The importance of understanding AI and how to interact with it, compared to knowing how to use Google effectively.

Aram B's advice to get hands-on experience with AI tools to leverage their benefits.

The limitations of codeless automation tools and the need for professional coding skills for complex scenarios.

The prediction that AI will replace coders but not engineers, as engineers connect the pieces of the puzzle.

The necessity of learning the basics of coding to understand and connect code pieces effectively.

The potential issue of AI hallucinations, where AI provides incorrect solutions that seem right.

The importance of human oversight in AI automation to catch false positives or negatives.

The incident with CrowdStrike's cybersecurity software, where AI automation reported a false positive.

The upcoming feature of 'self-healing tests' in AI-powered automation tools and the questions it raises.

Elon Musk's approach to training AI with a cutoff time to ensure the data's relevance and accuracy.

The challenge of AI training on false data and the impact on the results it produces.

The humorous observation that AI can become 'stupid' by learning from humans' less intelligent queries.

Transcripts

um hey everyone this is Alex USA days

and uh today I'm with Aram B um he is a

QA instructor and I actually uh took one

of his courses on becoming sdat with

playright uh very good quality of

recording very uh easily explained and

understandable easy to follow along

there are slides so if you're looking

for a course and you want to be as that

so software developer intest using

playright one of the most modern tools

for automation go ahead and check out

his uh web page it's uh bonder

academy.com and I will have a link in

the description uh there is a pretty uh

affordable plan like for there there's a

there's a plan for 100 bucks you can buy

the whole course and you can from start

to end go through the whole thing

unlimited time given so I think it's

very comparable to resources like you to

me but a lot a lot more uh a lot more

support so there's chat there's slides

there's like a lot more content uh that

you can work with in much better format

thank you Alex thank you for joining uh

and inviting me for this nice talk this

today what's Thursday evening yep and uh

I would love to talk yeah so uh one of

the things that I've been asked a lot

and I kind of shared my opinion a little

bit on it but maybe not you know

extensively enough uh people are

concerned about

Ai and uh they're asking if AI will take

their job specifically for qual

Assurance engineers and what can they do

if it will take their jobs so they can

prevent that um I don't know if you've

seen any Trends on AI tools doing

automation or anything like coming up

that it is possible that the QA jobs are

going to be taken by I what do you think

about that oh yeah I'm monitoring this

this subject pretty closely and I just

recently recorded a new course uh on my

platform which called mastering test

trigger and test trigger is AI based uh

platform for the test automation it's

pretty pretty cool one when just writing

the prompts you can automate the

scenarios without a single line of code

so very interesting concept and uh there

are more and more companies currently

heads down building the tools like that

I was recently contacted by other

startup based in China they were looking

kind of influencers or something you

know people who will promote uh this

tool to to the audiences so they are

kind of ready to go to the market very

soon I think we'll see more and more of

that stuff

um more of those tools with those

capabilities but answering your

questions here's here are my thoughts

that um Engineers should not be afraid

of AI stealing their jobs they should be

maybe kind of afraid of people who use

AI who can steal their jobs because AI

by itself is just a tool uh what can who

can get the competitive advantage of the

market are people who leveraging AI

capabilities can boost their

effectiveness productivity

and so on so those kind of things and uh

I I would say Don't Panic about oh my

gosh AI is everywhere and and uh we're

going to lose the job in like five years

or so uh what what you have to do is to

uh learn how to use the AI I would

compare it similarly like uh when Google

uh showed up right on the market and uh

there are people in even even today

there are people who know how to Google

well and can quickly find the

information in Google and those who are

just struggling to pull something out of

the internet so it's something with AI

something like that so just get your

feet wet with what is AI so how to

interact with this system try different

Eyes by the way CH GPT uh perplexity uh

Cloud AI Gemini you name it uh and try

to begin interaction with this system

and see how you feel uh that you know uh

moment of working with this system how

helpful is this what are advantages and

disadvantages of the system you with the

experience you kind of find your uh pace

and your approach to AI how to use it to

leverage your particular skills so this

is what everyone should start doing if

you are not using AI yet this is the

first step you you should start doing

register account in open Ai and chat GPT

or whatever and try to leverage this

tool instead of Google try and something

with it and see what what it's going to

be I have some exposure to uh codess

automation tools not a lot but a little

bit and what I found uh through my

experience is that even if you have a

tool that creates code U for you you

kind of drag and drop pieces all it and

the logic uh it it is not flexible so

unless you really know code it is really

hard to create with those tools what you

need to because what you end up doing

you'll have some rigid tests but if

there is anything that you want to build

around them and have like multiple entry

points stuff like that uh navigating

from one page to another like complex

end to end flow you'll have to edit the

code and some tools that uh have this

codeless automation they will allow you

to go under the hood and kind of uh Twix

snares and others don't so I don't I

don't see it as much of a um a threat

maybe U it's a good start if you want to

get an automation just understand it but

if you want to become a professional

most likely you will still have to go

hands on coding to write automation

scripts absolutely absolutely and I

would I would tell you even

more that um AI eventually will and I I

remember some prediction on Twitter you

know I monitor that AI eventually from

iws as far as far as I remember that AI

will replace the coders but not

Engineers because Engineers essentially

are those who connecting those pieces of

the puzzle together to make it working

as a single single machine so AI

probably will be able to write the code

for you but to connect all those pieces

together working as a final product that

you have uh that that you need that will

do the stuff that you need this is what

engineer is going to do and to

understand how to do that you definitely

need to learn the basics uh and learning

the basics of the coding because you

have you know the 100 lines of code over

here but in order to connect 100 lines

here and here you you should at least be

able to read this code to find a way

okay how they I connect that and if

something goes wrong you should

understand that this the basic concept

if AI gives you the suggestion what you

what to do you should understand what is

it about without understanding the

context it will be just you know Wild

West and what it's even what what even

worse and Scar situation is um You

probably aware of AI hallucinations

right so when AI gives you the

information pretending that this is the

right solution but it's actually not and

only human is able to catch it hey dude

you are giving me some wrong stuff it's

it's wrong and they are like oh really

I'm sorry yeah you're right this is the

wrong solution here is the right

solution you will not be able to catch

those without understanding the basics

and knowing the context veryable so the

basics and context is still not going

anywhere in my opinion yeah um I've

noticed this a lot so I use AI to help

me code I use

co-pilot uh and it it is very helpful it

speeds up your coding tremendously you

can write code so fast a lot faster than

uh you would write on your own because

it gives you tons of good suggestions

that you don't have to go and search for

and you just press Tab and it you know

writes all the code that it uh suggests

but one problem with it uh many times

it's not aware of the project where

you're working in so it's pulling data

from somewhere creating like path

variables stuff like that that it's

trying to import for example that don't

even exist in your project so there's

like a lot of random stuff that you have

to watch out for uh and a lot of llm

models right so they they work on basis

of data that exists and how they worked

with it before if you introduce any new

data doesn't matter how well trained on

previous data it will cause some sort of

hallucination happening because it will

disrupt the whole process flow of what

it has in place it will introduce this

variable and then it will start thinking

okay maybe I'll I'll try you know

elephant I've never seen an elephant

maybe it's green let me try it being

green so it's going to do things like

that um so from from what I've seen so

far you know you definitely want to get

into Ai and understand it and work with

it a little bit but in terms of

replacing testers or developers we're

not there yet and probably never going

to be there um one of the cases that was

recently I don't know if you have heard

of outage that was um caused by crowd

strike the cyber security software so

one of the things that if you read their

analysis on why it happened one of the

reasons was that they had their own

system Automation in place running tests

and it reported a false positive so

essentially the tests were crashing but

it says pass and that's how it the whole

thing went through the system without

someone looking at it open up those

tests and looking inside understanding

it's actual pass or something did fail

it was a false pass so if you leverage

everything to Automation and you give

everything away what what what is going

to happen if there's a bog in automation

itself it is still piece of software who

is controlling that who is going to

oversee if who going to fix your let's

not even go into automation let's talk

about development environment like ID

what if your VSS code does some weird um

I don't know adjustments of your

variables or data types for typescript

something and it saves like this and it

breaks everything like if no one's

looking if there's no real person

launching it running it and um verifying

that everything's correct you're just

going to have eventually going to have

an issue that's uh could be really you

know like a disaster that happened with

uh crowd strike right right and you know

what's interesting so when I reading the

the bullet points about some of those AI

powered automation tools one of the

advantages uh companies present about

the upcoming feature at least a

self-healing test you know and I'm like

I wonder how it's going to work how it's

figure out what self-heal and what is

actually a bug you know this is kind of

very

edgy uh kind of feature that it it can

self-heal failing test which actually

can be a bug right maybe it's going to

self-heal around an issue a bug until

the test passes but the actual issue is

still there exactly exactly I really

interested to dive and test this in to

when I will have the chance you know

with any of those tools to really test

this selfhealing feature so far I didn't

and one more thing by the way you you

you brought a good attention about the

data that was trained the AI is trained

on right and if it's not aware of the

data I recently um watched a interview

with Elon Musk and he was talking about

their xai U data center that they are

building and uh running their own Ai and

they told that they have um in in

algorithm training their data they have

a cut off time uh which data to use to

train not so they use the data before AI

was created because the date when AI was

created and a lot of that autogenerated

data just flown into the internet

because generate the text is not a

problem anymore with chat GPT but how to

determine the relevance what is true and

what is not so if you train AI on the uh

false data the AI will return you a

false result that's also a problem for

the researchers actually where's def

find a good data that AI can be trained

on yeah so what they do in their models

they create a cut of time before like

January 2022 or something and train

everything what is before that because

they know that that's the real data that

people created other than that later

than that it's kind of not sure I think

one of the common complaints that you

see when Chad GPT uh is released it's

pretty smart but over time from

interacting with people it becomes

pretty stupid it's funny so it trains

itself on the conversations with humans

and humans are asking different stupid

things and then all of a sudden your

Chad GPT is also doing stupid things

which is exactly funny exactly