I’ve been in QA long enough to notice a pattern. When people say testing, they often mean finding bugs.
But that’s not how it is in real life. In my book "Navigating Testing and Quality Assurance" a.co/d/0fPHT3T, I wrote my own definition based on what I kept seeing in projects:
"As I see it, testing is more than just finding bugs, it is a process of improving product quality, ensuring compliance with requirements and regulations, and delivering a product that meets or exceeds user expectations."
That didn’t come from a textbook. That came from shipping software and seeing what actually matters.
Because the best testing work I’ve done wasn’t just logging defects, it was:
- stopping failures before they hit production
- surfacing risks early (before they become expensive surprises)
Ilya Sutskever, one of the people who proved that “just scale it” works in AI, now says the scaling era is over. Current models are insanely good on benchmarks but generalize far worse than humans, which hints we hit an architectural ceiling, not just a compute limit. His new $3B company, SSI, is betting that 5–20 years from now, human-like learners will come from new research ideas, not the biggest GPU cluster.
Need more vibe coding angry birds flying in the picture. Also AI should be inserted under aws, cloudflare, azure with all the outages in last two months.
If 2025 is the year of AI agents, 2026 will belong to agent managers—humans who direct teams of AI workers.
I tried managing 4 AI agents… and it broke my brain. They constantly ask for clarification, need permissions, launch web searches—and I have to babysit all of it.
Sometimes they finish a task in 30 seconds. Other times, it’s 30 minutes of chaos. Half the work ends up in the trash because they misinterpret simple instructions.
Managing AI isn’t automated. It’s overhead.
What happens when you’re managing 10? Or 100?
Agent management might be the next job category—and the next burnout trend.
Have you tried running multiple agents at once? How many before you hit your limit?
alexusadays
alexusadays
I’ve been in QA long enough to notice a pattern. When people say testing, they often mean finding bugs.
But that’s not how it is in real life. In my book "Navigating Testing and Quality Assurance" a.co/d/0fPHT3T, I wrote my own definition based on what I kept seeing in projects:
"As I see it, testing is more than just finding bugs, it is a process of improving product quality, ensuring compliance with requirements and regulations, and delivering a product that meets or exceeds user expectations."
That didn’t come from a textbook. That came from shipping software and seeing what actually matters.
Because the best testing work I’ve done wasn’t just logging defects, it was:
- stopping failures before they hit production
- surfacing risks early (before they become expensive surprises)
- confirming we’re meeting requirements + compliance constraints
- protecting user trust
Then recently I read "Taking Testing Seriously" a.co/d/3J7J47g and this line hit me:
"Testing is the process of evaluating a product by learning about it through experiencing, exploring, and experimenting with it."
And I was like… yep. That’s the missing piece.
My definition is the WHY.
That quote is the HOW.
Put them together and you get the full picture:
WHY: quality, compliance, user expectations, confidence
HOW: learn the product by exploring, experimenting, and experiencing it
That’s the mindset shift I wish more teams made:
Testing isn’t only about the "bug hunt".
Testing is risk-driven learning that helps teams ship quality products.
How do you define testing in one sentence?
#softwaretesting #qualityassurance #qa #testing
3 days ago (edited) | [YT] | 14
View 2 replies
alexusadays
I must say I enjoy this thumbnail a bit too much 🤣
https://www.youtube.com/watch?v=eDh92...
2 weeks ago | [YT] | 7
View 0 replies
alexusadays
Ilya Sutskever, one of the people who proved that “just scale it” works in AI, now says the scaling era is over. Current models are insanely good on benchmarks but generalize far worse than humans, which hints we hit an architectural ceiling, not just a compute limit. His new $3B company, SSI, is betting that 5–20 years from now, human-like learners will come from new research ideas, not the biggest GPU cluster.
#AI #AGI #AIResearch #ScalingLaws #FutureOfAI #IlyaSutskever #SafeSuperintelligence
Source: www.implicator.ai/ilya-sutskever-declares-the-scal…
4 weeks ago | [YT] | 2
View 0 replies
alexusadays
Need more vibe coding angry birds flying in the picture. Also AI should be inserted under aws, cloudflare, azure with all the outages in last two months.
#ai #testing #qa
1 month ago | [YT] | 7
View 0 replies
alexusadays
AI reshapes support, but humans stay central—and organizations that ignore that reality will have to walk back their plans.
1 month ago | [YT] | 5
View 0 replies
alexusadays
AI is finally making a difference. We get a major outage on a weekly basis now 😂
#ai #outage #microsoft #amazon #qualityassurance
1 month ago (edited) | [YT] | 5
View 0 replies
alexusadays
My audiobook “Navigating Testing & Quality Assurance” is now available on Audible! Check it out 😊
www.audible.com/pd/B0FM4LT5JM?source_code=ASSORAP0…
4 months ago | [YT] | 12
View 0 replies
alexusadays
Flash sale on my new Udemy Playwright with Typescript course: www.udemy.com/course/playwright-automation-with-ty…
This will only last few days. So get it now :)
#alexusadays #playwright #qualityassurance
4 months ago | [YT] | 6
View 0 replies
alexusadays
I asked hashtag#ChatGPT to generate USA themed alphabet with image representing each letter, here are results:
4 months ago | [YT] | 4
View 1 reply
alexusadays
The Rise of the Agent Manager the new jobs ahead.
If 2025 is the year of AI agents, 2026 will belong to agent managers—humans who direct teams of AI workers.
I tried managing 4 AI agents… and it broke my brain.
They constantly ask for clarification, need permissions, launch web searches—and I have to babysit all of it.
Sometimes they finish a task in 30 seconds.
Other times, it’s 30 minutes of chaos.
Half the work ends up in the trash because they misinterpret simple instructions.
Managing AI isn’t automated. It’s overhead.
What happens when you’re managing 10? Or 100?
Agent management might be the next job category—and the next burnout trend.
Have you tried running multiple agents at once? How many before you hit your limit?
#AI #AgenticAI #FutureOfWork
5 months ago | [YT] | 5
View 0 replies
Load more