I tried Amazon's Dash cart while grocery shopping. I probably won't again.

I visited the Amazon Fresh grocery store in Fairfax, Virginia, to test a Dash cart.

This is one of 28 Amazon Fresh stores where you can use a Dash cart to shop, per Amazon's website. The company also offers them at a few Whole Foods stores and a regional grocer in the Midwest.

The Dash carts were on display as soon as I walked into the store.

As soon as I walked up to one of the carts, an employee greeted me and asked: "Would you like to try this new technology?"

He then showed me how to start using the cart. You have to download the Amazon Shopping app to your smartphone, log into your account, and then pull up an "in-store code" to scan using one of the cart's three scanners. That way, anything you place in the cart can be charged to your account.

I saw even more Dash carts as I headed toward the produce section.

This location had plenty of Dash carts — almost 20. During my visit, I only saw two other shoppers using them, while the others in the store were using baskets or regular carts to do their shopping.

The Dash cart has lots of on-screen features, like a shopping list.

This shopping list allows you to add items manually or through Alexa, Amazon's voice-activated assistant.

The cart also had videos showing you how to use it.

Amazon has added instructional videos to the Dash cart that show you how to use it. I watched this one, for example, which clarifies how to remove an item from the cart once you've scanned it.

I headed to the produce section to make my first selection with the Dash cart.

Produce was a good test of the Dash cart since most items don't have barcodes, and many are sold by the pound, making accurate weighing key.

That meant I had to enter PLU codes — those numbers on the stickers you find on produce — before placing it into the Dash cart's basket and weighing it using the built-in scale.

These bell peppers were priced per pepper, but I still had to find them on the Dash cart's screen before placing them in its basket.

Entering codes for produce reminded me of using self-checkout at other grocery stores.

I use self-checkout a lot at my local supermarket. Punching in codes for the bell peppers and other produce reminded me of that experience — and made the Dash cart feel like a self-checkout on wheels.

Dash carts also have a "sales & deals" tab that changes as you move around the store.

This selection of cranberry juice specials, for instance, popped up as I walked into the aisle where the products were located. The same thing happened later as I walked down the snack aisle and past the cooler case for milk.

Most products could be scanned via barcode, though some were easier than others.

The Dash cart includes three scanners: Two front-facing ones above the basket and another downward-facing one under the handle. You could scan most of the packaged goods in the store using one of the three sensors.

The position of the barcode on some products made scanning difficult, though. This four-pack of tonic water had its barcode on the bottom of the cardboard holder, making it tough to scan while keeping the glass bottles from slipping out.

I didn't get any error messages when I rearranged items in the cart or put items that were already there in my shopping bag.

Unlike many self-checkouts I've used, which give you an error message when you rearrange stuff in your bag or knock something over, the Dash cart didn't mind when I had to rearrange my purchases. I also didn't encounter any problems consolidating things into my shopping bag.

Removing items required using the screen but was easy to do.

While looking at the alcohol selection, I decided to see how hard it was to remove something from the cart and my bill.

Initially, I chose a can of hard seltzer, scanned it, and placed it in the Dash cart. Then, I picked it up out of the cart and put it back in the cooler.

The Dash cart asked me to confirm whether I had actually removed a product or not.

This screen popped up when I returned the can to the cooler. I confirmed that I had indeed removed it and proceeded to add a different choice to my cart.

The Dash cart didn't catch everything, though.

Given that the cart relies on its internal scale so much, I wanted to see what would happen if there was a discrepancy between the items on the screen and what was actually in the basket. (No, I did not steal anything, and neither should you.)

I scanned an item, then removed it from the Dash cart.

I started by scanning this box of mac and cheese and placing it in the basket. Then, I put it back on the shelf.

When this message flashed on the screen, I said that I hadn't removed anything.

I was still getting charged for the mac and cheese even though it wasn't in my cart anymore. Once I hit "No, my receipt is correct," I could keep shopping, and neither the cart nor the employee working the exit for Dash cart users caught it.

This is a big difference from Just Walk Out, which Amazon touted as capable of catching whatever customers walk out with — down to whether the avocado they picked up was organic or not.

Buying a hot cup of coffee also presented a challenge for the Dash cart.

Despite my last sub-par experience with Amazon Fresh's coffee, I decided to pick up a cup just to see how the Dash cart would handle it.

Placing the coffee in the cart's cup holder didn't work.

There was no barcode to scan on the cup, so I entered a PLU number on the screen and set it in the cup holder. After all, what if I wanted to sip as I did my shopping?

But the Dash cart's scale didn't appear to extend to that part of the cart, since it didn't charge me for it.

Eventually, I got the Dash cart to add the coffee to my receipt by placing it in the cart itself.

I felt like the hot coffee might spill on my shoes and the floor if I moved too quickly. Fortunately, I was near the end of my shopping trip.

When it was time to leave, I headed toward this special lane for Dash cart users.

An employee greeted me as I pulled into the lane. To my left were regular self-checkout options for customers using traditional shopping carts.

On the Dash cart, the screen prompted the employee to check my ID since I was buying alcohol.

After a quick verification, the employee directed me toward the entrance. She didn't check any other part of my purchase and didn't catch the box of mac and cheese that I had returned to the shelf but not removed from my receipt.

Amazon charged my account as I exited the Amazon Fresh store.

I got an email immediately after I went through the Dash cart lane confirming what I had purchased and charging my payment method on file with Amazon.

I had to leave the Dash cart in the store and carry my purchases out to my car.

As I left the store, an employee was waiting to claim my Dash cart. That meant I couldn't take the cart with me and load everything into my car.

Fortunately, my purchases all fit into a reusable shopping bag I brought. But if I had done a proper weekly shopping trip, I probably would've needed to either take the cart to my car or pull up to the store to load everything.

Overall, I'm not convinced that the Dash cart is the future of grocery shopping.

Using the Dash cart felt a lot like self-checkout: You have to scan everything yourself and punch in codes for some items. And, if you change your mind about something, you have to remove it from your receipt using the screen.

It's a far cry from the promise of Just Walk Out, which involves no scanning on the customer's part — other than to identify yourself when you enter the store.

My experiment with the box of mac and cheese also shows that, just like self-checkout, Dash carts aren't foolproof. Many retailers are pulling back on self-checkout because of theft, and it seems like the Dash cart has similar vulnerabilities.

Do you work at Amazon, Amazon Fresh, or Whole Foods and have a story idea to share? Reach out to this reporter at abitter@businessinsider.com