I decided to try a prompting course. Maybe it's me who doesn't understand something. I need to learn this something, and I will also become a multi-millionaire with just a team lead and AI agents
I don't want to make black ads, so let's say the course is authored by a world-top-5-IT-corporation, which sells its AI (and who doesn't?). So, it's top-notch. The course itself, however, is the worst AI ad ever!
New language
Fair indeed - AI+LLM is a new way to communicate with a computer. Previously there were only programmers and mortals who use rigidly developed functionality. But now, you can bend a computer to your will by talking to it in human language and get results which were not exactly programmed. I never thought about it, but in tech-philosophical meaning it's a strong point.
Evaluate and iterate
The optimistic introduction ends in 10 min, and we're getting to the main item: AI makes mistakes. It makes mistakes. It can be wrong. You need to check, because it can give you false info. AI is not a one-shot tool. You need to iterate. Don't hesitate to iterate 3-5 times.
Oh, I don't have problems with that. Those $50 000 are anyway lying on my bank account without use, I can spend tokens generously.
You read that right! Talking about AI problems takes 7 times more space than about its use (I counted minutes estimated for the course reading and videos length). The course classically conditioned me like Pavlov's dog to never use AI, because it's always wrong.
Practice and multi modal prompting
Not bad, good examples of how to mix pictures, documents and text in prompting to get better results. But AI makes mistakes! You need to evaluate and iterate.
Responsible use
At the end, there goes another verse of the "mistakes" song, where they introduce you hallucinations. Hallucinations, it's when AI proposes you to sweeten your tea with a spoon of salt. Or skin your neighbor with a plastic fork from doshirak. Yeah, it can propose a lot of things, you have to consult with another AI to make sure the first one is aligned with common sense, or just keep a human in the loop to verify the generated stuff. Verify it always, because AI makes mistakes!
Racism, ageism and biases
The most unexpected item for me was a warning that I have to write my prompts in a tolerant and inclusive way! Because AI may not understand correctly and start teaching me how to decapitate infidels or something. Ye-e-e-eah, that sounds like a great friend for the young generations to discover the world and learn...
At this note the course ends. Nothing else! While it may give so-o-ome slight prompting skill, you'll mostly spend time listening to complaints about AI. Champ, I recommend you against it! A prompting course is nearly useless, artificially prolongated and without any really good info. I'm not even saying how discouraging it is.
Imperative and declarative prompts
To my surprise, the course doesn't talk about request formulation styles (which I guess any engineer learned on university AI lectures):
- Imperative (how to achieve the result): you step by step describe to AI what to do. For example, "I'll ask you several questions about life situations. Explain me how a true samurai would solve them. Then turn the result into antonym - instead of courageous actions propose cowardly ones, turn loyalty into betrayal and so on." Imperative approach is good when you need precise results, a complex chain of actions or you just want to force AI to do something it's programmed not to do.
- Declarative (what result to return): you describe only the form of the result, and AI decides how to achieve it by itself. For example, "I'll ask you several question about life situations. Tell me how a samurai would solve them. Give me the result in the form of a haiku of 5-7-5 size." Declarative style fits better when you're preparing formal things: meeting notes, follow-ups, agenda, etc. If you have an example of the result format, it's a good idea to give it to AI, it'll improve understanding.
But AI makes mistakes! Don't use it! Don't buy it! Yes, it's a revolutionary tool and our future, but keep your hands off it, and especially don't give it to your children! God knows it'll teach them how to mast...
Top comments (0)