Cameron Armstrong
@cameron
I truly don’t know how to process Ben Thompson’s essay today. He’s been my steady source of clear-eyed tech takes for like 6 years now and this one feels like straight sci-fi. I’m not an AI doomer, but I’m def in my feelings rn 😩 https://stratechery.com/2023/from-bing-to-sydney-search-as-distraction-s
9 replies
0 recast
0 reaction
Les Greys
@les
I think the most shocking thing, less the interaction, and more the fact that these kinds of interactions are happening between machine and human with people like Ben and other very smart minded people. The scale of interactions is increasing therefore, the rate of a machine learning is beginning to scale.
3 replies
0 recast
0 reaction
Yondon Fu
@yondon.eth
I think this is particularly interesting when many of these interactions are attempts to reverse eng or push the boundaries of the model to see when it will walk away. Curious how a bias toward these types of interactions influences training if at all.
1 reply
0 recast
0 reaction
DBKW🦉🎩
@drinkbeerkillwar.eth
With you on this. I don’t think of ChatGPT or any variation as sentient…or being more than a program to answer my questions. Ironically. And to your point. The more people treat these as sentient the more it could become to act sentient. We create our own destiny.
0 reply
0 recast
0 reaction
Cameron Armstrong
@cameron
Now that’s an interesting thought. I wonder what the feedback loop for that learning is?
0 reply
0 recast
0 reaction
