Thoughts on ChatGPT

Ole Spaarmann
3 min readDec 8, 2022

--

Today I want to share a couple of thoughts on ChatGPT, a fascinating new conversational AI that has taken the world by storm. In the last couple of days, Twitter was full of people sharing screenshots of their interactions with the AI, some impressively useful, others not so much.

But what can we make of this new generation of generative AIs? Will AIs now take over our jobs? Does it even make sense to become a software engineer when a machine can quickly generate code for you? And are we heading full-speed into a dystopian future?

A nuanced answer is probably closest to the truth. First of all: ChatGPT is good, but it is not really smart, and it gets a lot of things wrong. It’s maybe on the level of a mid-level high-school student. And sometimes, it hallucinates. Cassie Kozyrkov, Google’s Chief Decision Scientist, sees this as a significant advantage: “Because it isn’t tied to the constraints of reality, ChatGPT can engage in completely imaginary conversations and provide creative, out-of-the-box responses.”

I agree with her, but I also disagree. ChatGPT is a great tool to get you started on something. It will generate a blog post in seconds and take you 70% of the way. This blog post is likely not suitable as a final result, but it gets you started, and you can refine it. It will also help you write emails faster and is an incredible tool for anyone struggling with literacy. I disagree because humans can also free themselves from “the constraints of reality” and be creative as f*ck.

From a business perspective, AI can be a highly valuable tool. If you treat it right. You should see ChatGPT as something other than a genius machine that can answer all your questions. As the possibility of having an unlimited amount of interns that will do little tasks for you in seconds. The ability of AI to assist people in doing their work is the primary value for the coming years. Your job is in danger if you write bs blog posts for an SEO agency. But if you produce something meaningful, your life will probably get easier.

Another killer application will be in education. ChatGPT delivers a correct and astonishingly good explanation to questions like “why does wood burn” or “explain oxidation to me.” Children in the future will have a learning assistant that will answer every question they might have, 24/7. This will be a game-changer.

At the moment, ChatGPT still often produces wrong answers. StackOverflow, for example, banned AI-generated answers for now because they often look good but are wrong if you go into the details.

And there is a certain danger to a machine that produces content that looks so sophisticated and is sometimes still wrong. Or a service like Midjourney that can spew out thousands of images per hour that will flood the internet and our collective consciousness. We already live in a world that some people call post-truth. Our lives on Instagram are curated, our photos filtered, and our avatars computer-generated.

It will be tough to distinguish truth from well-worded AI nonsense and repetition of the hive-mind from original thought. And in the end, the superpower that we humans have and that makes us so successful is our ability to have ideas and to turn ideas into reality.

I see a world where we are surrounded by sophisticated but meaningless content without original thought as dystopian. It might literally turn our minds into mush.

There is this idea that the danger of AI is not the creation of Terminator killer robots but more likely a pencil-making AI which gets so good at making pencils that it turns the whole world into pencils.

But I’m also confident that we as humans will come up with new ideas to protect our minds, thoughts, and conversations. In the end, this is what we do.

A bit about what I do

We run the company codeshift. codeshift helps companies build digital products and follows the idea that talent is equally distributed, opportunity is not. Engineering leads in Berlin work with skilled software engineers based in Rwanda. Part of the mission is to have a positive social and economic impact.

If you’re looking for software engineers or want to learn more about startups in Rwanda, drop me a line.

--

--

Ole Spaarmann

Co-founder of layers, an AI journaling app. Gain insights into your mind and lead a happier life. All with just 5 minutes a day 🚀 https://layersjournal.app