Close Menu
Webmemo Consulting Walter Schärer | Generative AI Beratung
  • Home
  • Social Media
  • Online-Tools
X (Twitter) LinkedIn Facebook RSS
  • Home
  • Kontakt / Über mich
  • Datenschutz
  • Nach oben
Webmemo Consulting Walter Schärer | Generative AI Beratung
LinkedIn YouTube RSS
  • Home
  • Social Media
  • Online-Tools
Webmemo Consulting Walter Schärer | Generative AI Beratung
NavigationHome»Chatbots»Let ChatGPT chatbot code your 3D space
Cube array visualized in Unity based on ChatGPT instruction
Unity cube array based on ChatGPT instruction
Chatbots

Let ChatGPT chatbot code your 3D space

Walter SchärerBy Walter Schärer17. September 20233 Mins Read
Hier teilen
LinkedIn Twitter Facebook Email

ChatGPT by Open AI is an amazing chat bot. It can even come up with instructions and code for 3D tools like Blender, Unreal Engine or Unity.

Chatbot ChatGPT by Open AI for coding

Open AI’s ChatGPT is all over the place since its release 3.5 in late 2022. «GPT» stands for its «generative, pre-trained transformer» algorithm.

For my first test, I go all out and task the chatbot with providing me with a Python coding script for the 3D modeling tool Blender that creates an array of cubes in space.

Here’s the prompt I feed the chatbot:


ChatGPT prompt asking for a Python script for Blender.
Prompt for ChatGPT asking it for a Python script


And this is the chatbot’s response, including actual code that I copy and paste into Blender:


ChatGPT's Python code for Blender
ChatGPT's Python code including instructions how to use the code


Once the script is copied into Blender’s scripting module, the renderer produces the desired array of cubes.


Cube array rendered in Blender based on ChatGPT code
Cube array as rendered in Blender


When I ask the chatbot to come up with the same for the 3D tool Unity, it produces the respective instruction including the C# script:


Instruction and C# script to build an array of cubes in Unity
Instruction and C# script to build an array of cubes in Unity


Which produces the respective visualization.

Below is Google’s «competing» result list:


Cube array visualized in Unity based on ChatGPT instruction
Unity cube array based on ChatGPT instruction
Google result list showing less quality results than ChatGPT
Google's result list for the same briefing.


Now, this offers all sorts of shortcuts for different tasks, think writing and debugging code or much simpler things like writing articles, emails and what not.

Apparently, the more specific the instruction, the more impressive the results.

I suggest you test the chatbot in your domain of activity. I’m blown away!

In the future, instead of asking Google for an instruction, I will first ask ChatGPT.

Here’s another example of ChatGPT’s superior service level when trying to achieve something in a software tool:


Google SERP on how to write text in Blender with Python
Google's result list promises a lenghty search experience.

Google's LaMDA chatbot

It’s a shame that ChatGPT garners all the praise because allegedly Google offers an even stronger chatbot called LaMDA. 

Potentially they better release it into the wild because users are very flexible. When Google first published their search engine in 1996, it felt like two weeks before nobody was using Altavista anymore.

«Let me Google that for you» did an amazing job at showing everybody just how much better Google’s new search engine was.

So from now on we may as well «chat things up» instead of googling them.

Inventor's dilemma: Google's "Kodak-moment"

With ChatGPT entering into ever more fields of application, e.g. powering Microsoft’s Bing search engine, Google will need to make up their mind:

  • Stick with the «old» model of monetizing result lists and risk becoming obsolete
  • Embrace the chatbot principle with direct answers instead of result lists and find a new monetizing model that may not make up for today’s revenues any time soon

Here are a few reasons why Google is holding back their own LaMDA chatbot

And here’s a New York Times write-up on Google rating ChatGPT a «Code Red» as of December 2022.

And here are six startups, that may end the days of search engines as we know them and put chatbots in place instead. Chatbots look poised to become the standard interface for human-machine interactions.

Much like described in Neil Stephenson’s sci-fi novel Snow Crash. In which he famously coined the term «Metaverse».

Anleitung Convenience Review
Share. LinkedIn Twitter Facebook Email
Walter Schärer
  • Website
  • Facebook
  • X (Twitter)
  • LinkedIn

Walter Schärer ist ein Generative AI Marketing Manager und Solutions Architect und bloggt bei webmemo.ch über Trends in künstlicher Intelligenz KI.

Ähnliche Informationen

How to prompt videos in Runway ML or Kling AI

25. Oktober 2024

AI image generation with Flux and the ComfyUI visual user interface

2. Januar 2025

The Best Day and Time to Post on LinkedIn? It’s Personal—Forget the Averages

25. August 2024
Add A Comment
Leave A Reply Cancel Reply

Über Walter Schärer

Dies ist Walter Schärers Blog über Online-Themen wie Generative AI, Digital Marketing, Search Engine Optimierung (SEO), Content Marketing und Performance Marketing.

Walter Schärer arbeitet seit 1994 als Scrum Product Owner und Online-Manager im Web-Umfeld.

  • Von HTML / VRML kam er via
  • Powerpoint / Word zu
  • Confluence / Jira dann
  • Trello / Whiteboard (Edding 500) und organisiert sich aktuell mit
  • Asana / Google Drive, wenn er nicht gerade mit
  • ChatGPT / MidJourney oder
  • WordPress / Elementor experimentiert.

«Programmierung» begann er mit

  • NoCode von Make und
  • LowCode von ChatGPT, ging über zu tatsächlicher Programmierung in
  • Python auf Anaconda und dann in
  • Google Colab / Gemini / Claude Sonnet sowie
  • Visual Studio Code / GitHub Copilot

Die Inhalte dieses Blogs spiegeln meine persönliche Meinung und sind von keinem Arbeitgeber beeinflusst.

© 2010 - 2025 Webmemo Consulting | Walter Schärer
  • Home
  • Kontakt / Über mich
  • Datenschutz
  • Nach oben

Type above and press Enter to search. Press Esc to cancel.