In 5 states machines will grade kids essays. An NPR article lays out how robots will grade free-form human written essays. Standardized tests in America are almost 100 years old — AI advances are talking education to the next level. Or maybe AI tools are automating a 100-year-old process.
Will kids start writing essays to pass machine testing rather then sharing ideas — I think yes.
Will students become better writers — that will likely still depend on the teaching more than the testing.
The concept of robo grading essays is nothing new, the technology is now better and fooling the machines is harder.
AI and automation tools are entering every part of our lives and while I am passionate about building technology products, we can not except AI help without question and reflection.
Here are some questions that come to mind any time we start using “fancy new” technology to solve old problems.
Are we using AI to scale a lousy existing process?
When you want to automate a process, it’s tempting to copy the way things work now. Instead, look at how automation allows us to work in new ways.
When you open a door with your hands you push, pull or turn a knob. It’s a simple process. However, think about the last automatic door you walked through –did the door swing open? No, automatic doors slide or rotate.
Automatic doors don’t copy the design of manual doors; automatic doors work in a completely different way.
AI tools temp us to automate a broken educations system. Instead, we should look at how automation can allow us to rethink the flow of learning.
Are we looking to machines for core meaning?
As we look to machines to solve the hardest human challenges, they will become our partners in unexpected areas.
We will undoubtedly look to machines for meaning at some point. Have you every ask Google a serious medical question, or about history or about how to solve your problem or your pain?
We are moving to a place where we look for meaning through machine filters.
It’s important to remember that machines can seem like cold unbiased and purely logically but they are still made by humans, and our bias is built in.
AI tools offer great insights into our world, but we should be cautious to take shortcuts to critically thinking about what we should care about.
Can machines teach us what’s worth saying?
If you want to communicate you need to have the tools to do so, but not at the loss of having something to say.
I never thought of myself as a strong writer, yet I felt I had strong ideas. Both are needed.
How do we learn what’s worth saying? We have to have teachers that point us to conversations, guide us in how to find personal meaning, and push us beyond our current understanding of the world. It’s a big task, but that’s what the best teachers do, they share knowledge in the context of the human experience and help us find our own original ideas.
Machines can test how we write, and likely help us write better, but teachers will still need to show us what’s worth writing about.
Today’s kids will make tomorrows machines.
Today’s kids will not merely be Digital Natives, they will be AI Natives. They will build things with a new perspective of the world, which those of us born before the great AI technology integration will find hards to grasp. Robot testings humans is not good or bad, it’s just part of the world we live in.
What questions do you want AI Natives to ask the machines they make?
P.S. All grammar errors in the article are ironic as I used AI to help check the grammar.
NPR article — https://www.npr.org/2018/06/30/624373367/more-states-opting-to-robo-grade-student-essays-by-computer
A Short History of Standardized Tests — https://daily.jstor.org/short-history-standardized-tests/
Automatic doors — https://en.wikipedia.org/wiki/Automatic_door
Originally published at Ron Sparks.