Key highlights of today’s newsletter: A recent study by the Institute of Software, Chinese Academy of Sciences, Microsoft, and others, suggest that the performance of LLMs can be enhanced through emotional appeal. Examples include phrases like “This is very important to my career” and “Stay determined and keep moving forward”.
Ever since chain-of-thought prompting became de rigeur, I've concluded that the best thing for me to do is to operate as though the LLM is "thinking" or "alive", however you want to put it. Having an organic conversation yields amazing results like nothing else.
Black box indeed! I'm sure we will discover some startling emerging properties in the next few years.
Yeah, it's interesting. But prompting can also feel unnatural or forced to be honest. The idea is that we get to talk to our computers in 'natural language', yet, the emergence of prompt engineering seems to contradict with that.
We can't just ask for what we want, instead we have to resort to all kinds of clever tricks and phrases to summon the behavior we want.
True. I kind of thinking of this phase as me learning the language and less the other way around, but if you're somewhat clever, you can generally get what you're after.
I'm still in awe that I can cast magical spells now. People who talk about generative AI as though it's nothing special or just some "bells and whistles" aren't really seeing what I'm seeing.
Ever since chain-of-thought prompting became de rigeur, I've concluded that the best thing for me to do is to operate as though the LLM is "thinking" or "alive", however you want to put it. Having an organic conversation yields amazing results like nothing else.
Black box indeed! I'm sure we will discover some startling emerging properties in the next few years.
Yeah, it's interesting. But prompting can also feel unnatural or forced to be honest. The idea is that we get to talk to our computers in 'natural language', yet, the emergence of prompt engineering seems to contradict with that.
We can't just ask for what we want, instead we have to resort to all kinds of clever tricks and phrases to summon the behavior we want.
True. I kind of thinking of this phase as me learning the language and less the other way around, but if you're somewhat clever, you can generally get what you're after.
I'm still in awe that I can cast magical spells now. People who talk about generative AI as though it's nothing special or just some "bells and whistles" aren't really seeing what I'm seeing.
First we learned to write our search queries in a way that would work best with search engines.
Now we’re learning how to write our prompts in a way that works best with LLMs.
Which is OK, we’re learning how to use tools.
But it’s not the same as being understood.