Love the newsletter Rich! And since attention is the limited resource all writers contend for, your consolidated approach of issuing the newsletter weekly is wise. The world needs less noise, not more and sending it once on Sunday diminishes noise while demanding focused attention!
Btw just read your article on PayTM. 😳I turned down a job in 2021 to bring PayTM to the US. So glad I did!
Hi Sarah! Thank you so much for the kind words! It's comments like yours that keep me writing.
The poll was a success, and 89% of 18 votes all want a weekly version, so I'll keep it weekly! I agree we are all inundated with content that struggles for attention.
Wow! What a story about PayTM! That was a close call. Early in my banking career, I was crestfallen when I didn't get a job I desperately wanted. Six months later, the company closed, and I realized that I dodged a bullet. Life is only understood by looking backward.
Hi Manorainjan, yes, I reread that a few times as well and think that the comment was for the flowers themselves, not the associated plant, which clearly used those other colors.
Let's just say that it was a nice literary tool for me to base my commentary on, despite its lack of specificity.
In this case it was not a political statement and quite irrelevant for our lives, like the great majority of comments about art. "no animals where hurt." ;-)
But I'm pointing it out, because in our political sphere, in the hybrid war where we are all combatants, it is often the case, that pictures are combined with intentionally misleading descriptions. And in the most cases, the onlooker truly believes to see in the picture, what the description tells them to see. This is not a matter of intelligence but of awareness. One has to be critical and ask oneself: "Is it true? Do I really see this? What does the picture really show?
Same goes for statistics and graphs. They plot a graph and tell You what it shows. Actually they tell You what it means. And they choose the metrics according to the desired outcome. We all know this famous case of GDP-comparison between China and the USA. And we know how different it looks using Purchasing Power Parity. Using the PPP metric, China has overtaken the US 10 years earlier than it seems to be using only GDP.
LLMs like ChatGPT use the same trick on all possible levels. You point out a mistake and it replies to be "sorry". No, AIs can't be sorry. And no, it was not a "honest mistake". It was the programmed attempt to mislead You.
It says it will make an attempt to be better. No, AIs can not make attempts. It will do the same "mistake" again and again.
With words it paints the picture to be Your humble servant and to "try" the best it could in order to "help" You. Thus the snaky piece of code is made to lull You in confidence about its good intentions. But Code has no intentions. Programmers do and product managers do and shareholders do. The shareholders, the only mighty ones, have the worst possible intentions regarding the users of LLMs. They want the AIs to be the most sneaky and efficient propagandists ever. They are the encoded leads on the road to hell. You thought the good old fashioned facebook-algorithm was bad? "You ain't seen nothing yet!"
Couldn't agree more LLM's can't truly be "sorry." Yes without a doubt there is both good and bad with LLMs. Hopefully, we'll get more good than bad out of them; time will tell.
I call myself a spiritualist.
Even if the whole planet explodes, my eternal Soul will reincarnate elsewhere.
Maybe my next skin will be green, who knows? ;-)
Love the newsletter Rich! And since attention is the limited resource all writers contend for, your consolidated approach of issuing the newsletter weekly is wise. The world needs less noise, not more and sending it once on Sunday diminishes noise while demanding focused attention!
Btw just read your article on PayTM. 😳I turned down a job in 2021 to bring PayTM to the US. So glad I did!
Hi Sarah! Thank you so much for the kind words! It's comments like yours that keep me writing.
The poll was a success, and 89% of 18 votes all want a weekly version, so I'll keep it weekly! I agree we are all inundated with content that struggles for attention.
Wow! What a story about PayTM! That was a close call. Early in my banking career, I was crestfallen when I didn't get a job I desperately wanted. Six months later, the company closed, and I realized that I dodged a bullet. Life is only understood by looking backward.
"with three shades of yellow ‘and nothing else’."
I can make out two shades of blue and green, white and black, at least.
Hi Manorainjan, yes, I reread that a few times as well and think that the comment was for the flowers themselves, not the associated plant, which clearly used those other colors.
Let's just say that it was a nice literary tool for me to base my commentary on, despite its lack of specificity.
In this case it was not a political statement and quite irrelevant for our lives, like the great majority of comments about art. "no animals where hurt." ;-)
But I'm pointing it out, because in our political sphere, in the hybrid war where we are all combatants, it is often the case, that pictures are combined with intentionally misleading descriptions. And in the most cases, the onlooker truly believes to see in the picture, what the description tells them to see. This is not a matter of intelligence but of awareness. One has to be critical and ask oneself: "Is it true? Do I really see this? What does the picture really show?
Same goes for statistics and graphs. They plot a graph and tell You what it shows. Actually they tell You what it means. And they choose the metrics according to the desired outcome. We all know this famous case of GDP-comparison between China and the USA. And we know how different it looks using Purchasing Power Parity. Using the PPP metric, China has overtaken the US 10 years earlier than it seems to be using only GDP.
LLMs like ChatGPT use the same trick on all possible levels. You point out a mistake and it replies to be "sorry". No, AIs can't be sorry. And no, it was not a "honest mistake". It was the programmed attempt to mislead You.
It says it will make an attempt to be better. No, AIs can not make attempts. It will do the same "mistake" again and again.
With words it paints the picture to be Your humble servant and to "try" the best it could in order to "help" You. Thus the snaky piece of code is made to lull You in confidence about its good intentions. But Code has no intentions. Programmers do and product managers do and shareholders do. The shareholders, the only mighty ones, have the worst possible intentions regarding the users of LLMs. They want the AIs to be the most sneaky and efficient propagandists ever. They are the encoded leads on the road to hell. You thought the good old fashioned facebook-algorithm was bad? "You ain't seen nothing yet!"
https://twitter.com/TuckerCarlson/status/1758529993280205039
Couldn't agree more LLM's can't truly be "sorry." Yes without a doubt there is both good and bad with LLMs. Hopefully, we'll get more good than bad out of them; time will tell.
"Hopefully"? I do not see reason for hope here.
But one has to differentiate between the "West" and China.
In the west the AI will strangle us so cozy, that we will digging our own graves freely.
But in China it will boost overall efficiency and comfort.
It is like the different interpretations of the myth of the dragon.
https://www.instagram.com/p/C3SkjNaCBuH/
Hi Manorainjan! Oh you can just call me an optimist! There is always reason to hope....but yes lots of reasons to fear as well!