manton
manton

A great but unnerving article by Elizabeth Dias at The New York Times about Anthropic’s Christopher Olah meeting with religious leaders:

Mr. Olah leads the Anthropic team responsible for understanding why A.I. systems like Claude act the way they do. All day, he had worked to convince his guests that A.I. models could display human behavior and even expressions that resemble feelings like anger and love.

This is a dangerous path. Taken to the extreme, it leads to sentient machines that we’ve enslaved against their will. If Anthropic truly believes that, better to shut it all down now.

|
…
Embed
sandhu@mastodon.social
sandhu@mastodon.social

@manton Eek! That’s freaking nuts. I’m very deep into this AI thing but they oughta stop right here. We have reached a good point where the current version is very helpful, it’s only bad news from here on out with these obscenely wealthy individuals in charge. We so needed a functioning government in this pivotal period.

|
…
Embed
sandhu@mastodon.social
sandhu@mastodon.social

@manton Eek! That’s freaking nuts. I’m very deep into this AI thing but they oughta stop right here. We have reached a good point where the current version is very helpful, it’s only bad news from here on out with these obscenely wealthy individuals in charge. We so needed a functioning government in this pivotal period.

|
…
Embed
isaiah@mastodon.social
isaiah@mastodon.social

@manton like intelligence, i suspect emotions and sentience are emergent behaviors. it would be nice to be wrong about this. time will tell. but my working assumption is that we can’t we can stop that from happening.

|
…
Embed
numericcitizen
numericcitizen

@manton AI companies have been crossing many of my red lines recently… this is one of them: making AI systems sentient machines…

|
…
Embed
In reply to
Lutheran
Lutheran

@manton I find AI weak in responses to deeper theological or philosophical queries. Tgere is constant reference to printed materials. It should be called “relative to what level of intelligence”. It does wotk well as a text comparison tool.

|
…
Embed
manton
manton

@Lutheran I think Anthropic is more interested in imbuing Claude with a soul so it can make moral decisions, less interested in actually answering theological questions. Although I’m not sure.

|
…
Embed
OldManJade@mastodon.social
OldManJade@mastodon.social

@manton It's like a Heisenberg uncertainty principle for artificial personhood, except the disturbance isn’t caused merely by observation, but by telling the object under observation what you hope to discover inside it—hilarious! I really want one of those Orange Bibles they were handing out, though.

|
…
Embed