Over the past year, AI has become an essential tool in my daily work, simplifying tasks, and saving me time and headaches. Whether it’s copy editing, translating, programming, or video production, AI enables me to work more efficiently, having to do much less boring tasks myself. In this article, I’d like to share how I’ve been using different AI tools and platforms. (If you haven’t been sleeping for the last year, this may contain no surprises).
In the above video I’m using a HeyGen generated AI avatar (current version) to get over the awkwardness of having to talk about myself – in German – in the introduction to pitchingmasterclass.com
1. Copy Editing and Translations with ChatGPT
One of the most significant ways I’ve integrated AI is through ChatGPT, which I use daily for a variety of tasks. No surprise there. It serves as my go-to tool for copy editing, allowing me to refine content with greater speed and precision. I also use it to write boring-to-write copy – like this article.
When I’m dealing with multilingual content, ChatGPT helps ensure that specific meanings – especially those influenced by cultural nuances or figures of speech – are accurately conveyed between languages.
This becomes especially useful when I use tech jargon or idiomatic expressions that don’t always translates well using other AI translation services. Whether I’m refining technical writing or simplifying complex concepts, ChatGPT helps maintain both clarity and meaning across languages.
Below, an example of Kling AI generating videos out of images and Suno generating a pop tune to the lyrics that ChatGPT made as a summary of my “TL;DR” lesson of pitchingmasterclass.com
2. Programming & Scripting
For most of my needs, ChatGPT has replaced the need to sift through StackOverflow (RIP) or various forums for simple programming queries, providing quicker solutions to snippets of code or scripts – that actually work. e.g. creating widgets for web pages, conditional modal pop-ups, automation, etc. Mostly pedestrian stuff if you’re a pro dev, but I now get mundane shizzle done in seconds instead of minutes or hours. To illustrate, I now get production level results done faster than it would have taken me to formulate a sensible google search query, let alone the time it would take me to communicate to someone else what I need in text or speech.
This is the way – I don’t know a single (good) programmer who doesn’t use AI in any way or form these days.
I’ve also been experimenting with Claude AI for programming, but I need more practice before I feel comfortable using it for production. I’ve also seen insane things built with v0 by Vercel, so it’s also on my todo list of experimenting further with.
Below, and older example of AI generated copy text, voice, and moving image – to pimp the benefits of pitch training and pitching competence also for employees in companies for my pitchingmasterclass.com corporate edition:
3. AI-Assisted Transcriptions, Translations, and Dubbing
Working on my educational content, I rely on the built-in AI tools of DaVinci Resolve (fantastic piece of software), particularly the voice-to-text transcription feature. This feature provides a solid first draft of text for the content I’m going to be localizing. Although the transcripts aren’t 100% perfect, they save a tremendous amount of time compared to other means as no additional work is needed – it’s already there in the program where the original video content is edited.
Once I have a rough proofread version of the transcript, I run it through (local hero!) DeepL for translation into the target language. While DeepL is accurate, I still do a final proofread to ensure subtle details are somewhat correctly interpreted. ChatGPT is a great help when I encounter quirky sentences, idioms, or technical terms. It helps me refine these translations to preserve the original context, ensuring that the target audience grasps the intended meaning fully in their own language.
4. Voice and Video Content with ElevenLabs and HeyGen
Once the translated content is done and proof-read, I use ElevenLabs for voice dubbing. This AI voice tool allows me to generate a virtual voice of my own, also in other languages. The convenience of having an AI-generated version of my voice reduces the need for tedious manual recording and dubbing, while maintaining some level of authenticity.
Additionally, I use HeyGen to create video content with an AI-generated avatar of myself. I’ve found it useful for adding some humor, breaking up longer / more boring content. It’s a way to keep audiences engaged using cheap tricks – for now.
Below an older test video I made with a previous version of HeyGen to test its (insane) language translation capabilities – watch AI-me speak as good as flawless English, Greek, Arabic, Finish, Chinese, and Ukrainian:
5. Image Content Creation with Midjourney
For all my illustration needs, Midjourney has been a game-changer. Instead of spending hours searching through stock images or paying for licenses (although paying wasn’t really a pain, the time wasted searching was a cluster headache), I now can generate custom visuals for my keynotes and educational content that fit with the style and tone I’m aiming for on-demand in minutes. This also has the added benefit of knowing that these images are unique and haven’t appeared (and won’t appear in the future) anywhere else outside of my own branded content (which, if you’ve been using royalty-free subscription services like Storyblocks to source content, you’ll soon enough find out is a real and regular thing: I don’t know how many times I’ve had signature soundtracks or key visuals I’ve used in my content suddenly also appear in other companies’ content anywhere from online ad campaign videos on YouTube to print billboards.)
6. Music and Video Generation for Creative Content
For funsies, I’ve been experimenting with tools like Suno to generate music based on lyrics I generated using ChatGPT. This has been especially interesting for educational content, where I set core topics to music. I see a lot of potential for organizations to use music and song in their internal communications, whether for onboarding new employees, explaining codes of conduct, or even training on fire safety routines. It’s a memorable, engaging way to deliver obligatory messages that may be awkward and boring – but still need to be communicated effectively.
To master the songs coming out of Suno, I’ve been using the AI mastering service from Dolby on SoundCloud as it allows you to also pay per master, not only as a subscription. There are many other mastering services available, but most require a subscription.
I’ve also been playing with RunwayML since launch, which allows me to create AI-generated videos based on reference images I design using Midjourney or regenerate interesting new videos using simple reference videos shot on the phone. While still somewhat experimental, Gen-3 is quite impressive when it works out, it offers an exciting glimpse into the future of AI-driven media production. Even though it can feel hit-or-miss at times, the potential of tools like this is obvious, and they are rapidly improving. This is where the industry is going.
Another tool for videos from images or text to video I’ve seen a lot of interesting results from is Kling AI. It may be able to do things Gen-3 cannot or will not do. However, in my experience it takes longer (even on a paid plan) for the results to materialise than with Gen-3 turbo, and I haven’t had a project yet where I’ve had the patience or luxury of waiting.
Below is an example of what happened when I asked ChatGPT 4o to write me some rap lyrics describing what Lean Startup is all about. It spat the lyrics out in seconds. Then I asked Suno to set the lyrics to an east-coast hip-hop rap tune, and it too spat out the result in seconds. I then asked Midjourney v6.1 to generate me a reference image of Steve OG Blank that I imported into RunwayML Gen-3 turbo to generate moving videos from the reference image:
7. Corporate (Still) Missing the AI Train for Communications & Training?
One observation from my anecdotal experience is that many organizations aren’t really onboarded the idea of using AI for internal communications and training – yet. Tools like ChatGPT, Suno, HeyGen, and ElevenLabs transform how companies can generate content – especially for the extremely boring but highly necessary quadrant of topics, IMO. By setting key topics to song, or using AI avatars for onboarding and training, companies could make otherwise mundane information more engaging – not to mention the costs it would save AND the speed and costs of keeping the content current, improving incentives to actually release a new version AND the time from when need arises to actually updated content delivered .
I cannot stress enough how easy and fast this process is. Below an example of a country banjo tune lamenting the downfall of twitter after Elon Musk bought it and killed the blue bird – written by ChatGPT. This took only a few minutes to make in total. Now imagine it’s the code of conduct for your conference being summarized in a pop song instead:
Actually, you don’t have to imagine it. Here’s an example I cooked up (for educational purposes, no affiliation with the Black Hat conference whatsoever – obviously) asking ChatGPT to create a summary of the Black Hat hacker conference code of conduct as punk pop lyrics and Suno to create a punk pop tune to go with the lyrics.
It took me around 30 seconds in total:
Conclusion
AI has transformed the way I work. From copy editing and translating to voice dubbing and making content creation much faster and easier, especially for solo operators. These tools have allowed me to optimize my workflow to contain less boring and time-consuming tasks, allowing me to have more time for what I consider the more fun tasks of work – like not working at all. As AI continues to evolve, more organizations will discover the potential of these tools, or simply get left behind. Like always.