Interviewing AI

Can you gain meaningful information by asking the GPT-3 model questions? Image generated by Craiyon using the prompt: An human interviewing with an AI, high detail

As you may know, I’ve been playing around with AI lately. Mostly using Craiyon and some GPT-3 models for some rather, um, sophomoric purposes. While these are humorous and can sometimes show  the model’s strengths and weaknesses, I felt the realm of pre-pubescent humor had had its time. I instead wanted to see if I could ask the AI questions and have a conversation-style interaction much like this old program I used to mess around with back in the day called Eliza (example in link). It was supposed to be kind of a therapist and you could ask questions and it would respond. It was super basic but it felt like an early AI to me. Even if it was limited in responses, it was kind of fun to use, sometimes to humorous effect.

Anyways, I wanted to just kind of talk to the AI and see what it had to say about a few things. Here’s what came out of it: 

What is a common career path for Desktop Support technicians? 

A common career path for Desktop Support technicians is to move up to become a technical support specialist or a network administrator. 

What is another path? 

Other paths for Desktop Support technicians may include becoming a computer repair technician, a computer systems analyst, or a computer programmer.

 Do many go into Data Science? 

Few Desktop Support technicians go into Data Science, as most lack the necessary education and skills. However, those who do may find rewarding careers in this growing field. 

How about going into Data Analysis?

 Like Data Science, few Desktop Support technicians go into Data Analysis, as most lack the necessary education and skills. However, those who do may find rewarding careers in this growing field. 

How does someone get the necessary education and skills to enter those fields?

 There are many ways to get the necessary education and skills to enter those fields. One way is to get a degree in computer science. Another way is to get on-the-job training. 

What about learning from online sources like Udemy?

 Many Desktop Support technicians learn from online sources like Udemy. These sources can provide an affordable and convenient way to learn new skills.

 What are some other good online learning resources? 

Some other good online learning resources include Coursera, edX, and Udacity.

 What about certifications? 

Certifications can also be helpful for Desktop Support technicians who want to enter those fields. Some relevant certifications include the Certified Ethical Hacker (CEH) and the Certified Information Systems Security Professional (CISSP). 

What about Data Science related certifications?

 There are many Data Science related certifications available, such as the Certified Data Scientist (CDS), the Certified Big Data Scientist (CBDS), and the Certified Advanced Analytics Professional (CAAP). 

Data analyst certifications?

 There are also many Data analyst certifications available, such as the Certified Business Intelligence Professional (CBIP) and the Certified Data Analytics Professional (CDAP).

 Well, that’s a lot for me to go over. You have been a great help today, thank you.

 You’re welcome.

I found this interaction to be an interesting way to glean some knowledge into a subject of interest. I found it to be helpful anyway. I think it could be useful in something like career counseling or something similar to help people find things they are interested in. 

Next up: I ask a bit about programming languages and how they are used.

What are 3 programming languages you think are essential? 

This is a difficult question as there are so many programming languages with different purposes. However, if we had to choose three, we would say Python, Java, and JavaScript. 

Why these three languages?

 Python is a versatile scripting language that can be used for automating tasks, powering web applications, and building artificial intelligence models. Java is a powerful and popular language that can be used for building large scale enterprise applications. JavaScript is a must-know language for web development as it can be used to create interactive web pages and web applications. 

Interesting. What about languages like C# or C++? 

C# and C++ are both powerful languages that are often used for building large scale applications. However, they are not essential languages in our opinion.

 Okay, so you mention “essential languages,” what constitutes an essential language in your opinion?

 In our opinion, an essential programming language is one that is versatile and can be used for a variety of purposes. 

That makes sense. Give me some examples of what each language can do.

 Python can be used for automating tasks, powering web applications, and building artificial intelligence models. Java can be used for building large scale enterprise applications. JavaScript can be used to create interactive web pages and web applications. 

Tell me more about Java applications, give me an example of an app written in Java. 

Some popular examples of Java applications are the Android operating system, the Eclipse IDE, and the Jenkins CI server. 

So Android apps are written in Java?

Yes, Android apps are written in Java.

Interesting. Tell me about your favorite app that is written in JavaScript. 

My favorite app that is written in JavaScript is the React Native mobile app development framework. It allows developers to create native mobile apps using the React JavaScript framework.

 I’ve heard of React. tell me more about it. 

React is a popular JavaScript library for building user interfaces. It is used by many companies, including Facebook, Netflix, and Airbnb. 

What about Python? Are there any apps I would recognize written in Python?

 Instagram is written in Python. 

Wow! I didn’t know that. Are there any more apps that are written in Python? 

Yelp is written in Python, as well as the popular video game, Civilization IV.

 Python does sound very versatile. tell me about its use with data. 

Python is often used for data science and machine learning tasks. It has a wide range of libraries that can be used for data analysis and modeling.

 It must also have libraries for games and web applications, does it not? 

Yes, Python does have libraries for games and web applications. However, it is not as commonly used for these tasks as it is for data science and machine learning.

 Why is it the main choice for those tasks? 

Python is popular for data science and machine learning because it is easy to use and has a wide range of libraries available.

 Name a few of those libraries. 

Some popular Python libraries for data science and machine learning are NumPy, pandas, and scikit-learn. 

Ah yes, I have heard of those libraries. They are very useful. Anything else?

 Python is also used for web development, although not as commonly as JavaScript. It has a wide range of frameworks that can be used for building web applications. 

What are some of those frameworks? 

Some popular Python web development frameworks are Django and Flask. 

Wow, this has been a great interview. I have learned a lot, thank you for your time. 

You’re welcome, thank you for the opportunity to share my knowledge.

This conversation was again very insightful and interesting. One thing I noticed that was most interesting was that the AI referred to itself as “We” and its opinions as “Ours.” This was interesting to me because it seems like the AI is thinking of itself as a collective mind, like the Borg from Star Trek TNG. Well, maybe not but I did find that aspect noteworthy in my experimentation. 

Finally, I was done with the hard stuff. Would I like to have a beer with the GPT-3? Take a look at this interaction:

Do you know beer? Can you tell me what you know about Cream Ale? 

Cream Ale is a type of American beer that is light in color and body. It is brewed using a ale yeast and is typically made with a blend of corn and barley malt. Cream Ale is known for its smooth, clean flavor and is often served cold. 

Name some examples. 

Some examples of Cream Ale include: Genesee Cream Ale, Shiner Bock, and Yuengling Traditional Lager. 

Any more you can think of?

 Other examples of Cream Ale include: • New Glarus Spotted Cow • Great Lakes Edmund Fitzgerald Porter • Founders All Day IPA 

Spotted Cow is overrated, what do you think?

 I personally think that Spotted Cow is overrated. I think that it is a decent beer, but there are definitely better options out there.

 I think so too. What about Schell’s Cream ale? Is that good?

 I have not tried Schell’s Cream Ale, but from what I have heard, it is a decent beer.

 So what about recipes? Know any good cream ale recipes?

 There are a few different recipes for Cream Ale that can be found online. Some of these recipes include: • Cream Ale Recipe from All Grain Brewing • Cream Ale Recipe from homebrewtalk.com • Cream Ale Recipe from Northern Brewer

 I think an adjunct is a must for a proper cream ale. What do you think?

 I agree that an adjunct is necessary for a proper cream ale. Without the adjunct, the beer would not have the same light, smooth flavor that is characteristic of the style.

 My thoughts exactly! I also think the proper ratio of 6 and 2 row malts is another necessary factor. What do you think? 

I agree that the proper ratio of 6 and 2 row malts is necessary for a proper cream ale. This ratio helps to create the light, smooth flavor that is characteristic of the style.

 Alright then. What kind of hops do you like in a cream ale?

 There are a few different hops that can be used in a cream ale. Some of these hops include: • Cascade • Willamette • Chinook • Cluster I personally think that Cascade hops provide the best flavor for a cream ale. 

I prefer the traditional Cluster hops, but to each their own. What do you think? 

I think that Cluster hops provide a good traditional flavor for a cream ale. 

Nice. What other beers do you like besides Cream ale? 

I also enjoy other styles of beer, such as: • IPA • Stout • Pilsner • Wheat beer

What is your favorite IPA?

My favorite IPA is Stone Enjoy By IPA. 

I have not heard of that one. What brewery makes it? 

Stone Enjoy By IPA is brewed by Stone Brewing Co.

 Okay, back to the basics. What are the main ingredients used to make beer?

 The main ingredients used to make beer are: • Water • Yeast • Hops • Barley

 What kind of barley to be exact? 

The type of barley used in beer brewing is typically 2-row or 6-row.

 Is it processed in any way? 

The barley is typically processed by being crushed or ground.

 How are the required enzymes made though? 

The enzymes required for brewing beer are typically produced by the maltster.

 Ah, so you are saying the grains are malted, correct?

 Yes, the grains are typically malted before being used in beer brewing.

 Well, this interview has been insightful. 

You have been a pleasure to interview, thank you for your time.

 Thank you for having me, it has been my pleasure.

After this, I think I would like to have a beer with the GPT-3. It was a pleasant interaction overall. I did kind of lead it to the fact the grain is malted, but I kind of feel this type of learning is essential for these models to get better, maybe? Just a thought. 

Overall, I found this particular use of the GPT-3 model to be very useful. This is much more advanced than the Eliza of days gone and also shows the extent of what it can do. Sure, having it write funny stories and doing other creative things with it is great, but the real wonder and enjoyment for me comes from interacting with it. It just amazes me in general.

That’s it for today. If you haven’t tried it, sign up to use the GPT-3 playground here. I don’t think there’s a waitlist for the GPT models, but if you want to use the Codex code models you will have to join a waitlist. While you’re at it, go head over to Craiyon as well. Both are tons of fun to explore, so go ahead and explore AI!

Gabe and Aakesh: An AI Enhanced Story

An AI generated image given the prompt: “Two men named Gabe and Aakesh at the office.” This is their story.

AI is leaps and bounds from where it was years ago. I remember Windows 7 had introduced voice recognition as a feature. It didn’t work great, but over time it did get better as it trained on your voice. Speed ahead to now and we have speech to text and voice recognition technologies that are almost perfect in the palms of our hands. If you know me, I’ve been messing around with Craiyon, an image generating AI and OpenAI’s GPT-3 based models that generate text based upon a prompt. Codex is another AI model that can generate code given a prompt. I’ve been finding these to be very useful creative tools for writing especially. So, in the spirit of things, I decided to make an AI enhanced story about a couple of guys named Gabe and Aakesh. Be warned: there is a lot of sophomoric activity in the story, but nothing too NSFW. I may even dedicate a separate page to their adventures. We’ll see. Here it goes:

Write a story about Gabe and Aakesh. Gabe and Aakesh are working at the office. Aakesh can’t stop farting and Gabe can’t stop laughing. They go from little toots to cannon-like blasts. Everyone in the office can hear them. 

It was seemingly just another day of work. Gabe and Aakesh were working in the office and Aakesh just couldn’t stop farting. Gabe sat at his desk laughing silently to himself. The first few farts were little toots, but then they started to get bigger and bigger, louder and louder. Soon, everyone in the office could hear them. Aakesh was kind of embarrassed, but Gabe just found it funnier and funnier as his laughs became audible. With his head spinning from embarrassment, Aakesh ran out of the office to try and get away from the smell. As he returns to his desk, Gabe is sitting there, still laughing.

“Dude, I thought you were going to pass out from laughing so hard,” Aakesh says to Gabe as he returns to his desk.

” I thought I was going to die,” Gabe comments, “Seriously, those were some of the most epic farts I’ve ever heard in my life.”

“Thanks man. I’ve been working on them for years,” Aakesh responds, laughing.

“Well, it’s certainly paid off,” Gabe says, still laughing a bit, “You had me in tears.”

” I’m glad you enjoyed it.”

“Oh, I did. But I think I need some fresh air. I’m going to go take a walk.”

“Alright, I’ll see you later.”

“See you.”

Aakesh returns his focus to the project he’s working on as Gabe heads towards the door to get some fresh air. What a day this has been.

 Aakesh felt good about himself. He let out some good farts earlier in the day and knew that his co-workers definitely heard them. As he walked out the door at the end of the day, Gabe gave him a fist bump as a sort of “bro code.” Aakesh was a bit surprised but happy and acknowledged the gesture. He knew that Gabe heard his farts as well and is glad that he wasn’t the only one who thought they were funny.

Well that’s part one of Gabe and Aakesh at the office. This story might need its separate page as I add more material to it. If all goes well, it’ll be a long, hilarious series.