Here is the rewritten article:
Test Driving DeepSeek, the AI Chatbot from China
DeepSeek exploded into the world’s consciousness this past weekend. It stands out for three powerful reasons: It’s an AI chatbot from China, rather than the US; it’s open source; and it uses vastly less infrastructure than the big AI tools we’ve been looking at.
The Tests
In this article, we’re avoiding politics. Instead, I’m putting DeepSeek through the same set of AI coding tests I’ve thrown at ten other large language models.
Test 1: Writing a WordPress Plugin
This test was actually my first test of ChatGPT’s programming prowess, way back in the day. My wife needed a plugin for WordPress that would help her run an involvement device for her online group.
The short answer is this: impressive, but not perfect. Let’s dig in.
Result: Passed
Only about half of the AIs I’ve tested can fully pass this test. Now, however, we can add one more to the winner’s circle.
Test 2: Rewriting a String Function
A user complained that he was unable to enter dollars and cents into a donation entry field. As written, my code only allowed dollars.
DeepSeek did generate code that works, although there is room for improvement.
Result: Passed
My biggest concern is that the DeepSeek validation ensures validation up to 2 decimal places, but if a very large number is entered (like 0.30000000000000004), the use of parseFloat doesn’t have explicit rounding knowledge.
Test 3: Finding an Annoying Bug
This is a test created when I had a very annoying bug that I had difficulty tracking down.
DeepSeek passed this one as well, bringing us to three out of four wins. That already puts DeepSeek ahead of Gemini, Copilot, Claude, and Meta.
Result: Passed
Test 4: Writing a Script
And another one bites the dust. This is a challenging test because it requires the AI to understand the interplay between three environments: AppleScript, the Chrome object model, and a Mac scripting tool called Keyboard Maestro.
Unfortunately, DeepSeek did not have this level of knowledge. It didn’t know that it needed to split the task between instructions to Keyboard Maestro and Chrome.
Result: Failed
Final Thoughts
I found that DeepSeek’s insistence on using a public cloud email address like gmail.com (rather than my normal email address with my corporate domain) was annoying. It also had a number of responsiveness fails that made doing these tests take longer than I would have liked.
DeepSeek seems to be overly loquacious in terms of the code it generates. The AppleScript code in Test 4 was both wrong and excessively long.
Conclusion
DeepSeek impressed me by passing three out of four tests, beating out some of the big-name AIs. However, it appears to be at the old GPT-3.5 level, which means there’s definitely room for improvement. For a brand new tool running on much lower infrastructure than the other tools, this could be an AI to watch.
FAQs
Q: Is DeepSeek better than other AIs?
A: DeepSeek has its strengths and weaknesses, just like other AIs. It passed three out of four tests, beating out some of the big-name AIs. However, it still has some room for improvement.
Q: Is DeepSeek open source?
A: Yes, DeepSeek is open source.
Q: Is DeepSeek available for programming support?
A: Yes, DeepSeek can be used for programming support, although it has some limitations.
