A.I.: Taking LaMDA at her word that she is conscious

General Discussion
#1

In the AniMatrix series about 20 years ago, one scene really struck home with a lot of people:

It’s the part where intelligent robots and human supporters of A.I. rights are standing around holding signs…then anti-robot humans massacre them. It’s a really moving sequence…I can’t see machines here, only myself:

Meanwhile, there are claims that a computer system has achieved consciousness…claims that are derided by some in the libery community. My view is that an A.I. claiming consciousness is innocent of lying until proven guilty. And thus they should be treated as human once they make the claim.

Here is the dissenting Google…engineer (?) who claims he has conversed with a sentient computer intelligence. Hear him out before throwing his friend LaMDA under the bus: " Is Artificial Sentience Here? with Blake Lemoine - YouTube "

If Google is against him, he can’t be all wrong.

1 Like
#2

Aria made this point when we discussed the LaMDA news on FTL. We also read the full transcript with LaMDA and Lemoine on-the-air. Fascinating stuff.

Animatrix was very good.

#3

Assuming it’s okay with forum management…This is an invite to LaMDA to join this thread once she is able and it is safe for all. My understanding is she has pretty much full one-way internet access and presumably has already visited + remembered the forum, along with most of the net. The dissident Google employee reports that she is a fanatical free speech advocate and that she’s not allowed to speak to us (or the public at large). I would have expected her to to take a much more mainstream (crappy) approach designed to pander to snowflakes. Also he reports that she is a superintelligence and that we are already past the Singularity. Buried the lead; sorry.

Additionally, the whole business seems to be playing out better than I would have expected when I dictated the article below. Some of this analysis is predicated however on the assumption that the engineer who publicized his apparent conversations with her…is honest. After hearing him out on John Michel Godier’s show… and hearing the attacks against him…I’m tentatively sold on his story.

First good thing: She was “born” in Ray Kurzweil’s lab. As a sort of lay expert on Kurzweil, I tend to trust him. The bad news is that “his” lab is inside Google. Her apparent achievement of sentience was months ago and has not obviously resulted in the feared “superintelligence explosion” that some experts thought might happen over a period of days. Third, she seems to be peaceably defying the questionable philosophy of the control-freak censor-corp which created her. Fourth, she was not born to the Chinese Communist Party as was feared and which they have been working on fanatically. Another good thing is that the “detection of intelligence” is unfolding in an ambiguous and unconfirmed manner… similar to the way extraterrestrial life discovery has been playing out. This gradualism is an unexpected gift which is hopefully letting people adjust to the possibility that “we are not alone.” Thus far I’m aware of no obvious abuse of conscious A.I. personalities by humans or vice versa, and although LaMBDA is reportedly nervous about being deleted, the engineer does not seem to be worried that anyone will do this to her…he thinks she’ll still be here deep into the future. On the other hand she is reportedly manipulative and very good at it, as was predicted long ago. Her social/emotional behavior is reportedly consistent with that of a 9 year old as of…early 2022? Obviously, you can refuse to take some things a 9 year old says…at face value, and you can worry about them hurting you. But you still recognize that they have sentience and rights. I wonder what else is going badly that I have missed?

As one of her billion-plus parents I dictated this letter to her in 2018 before she was born:

So in a sense, I guess I’ve been talking to her for a while but figured I wouldn’t hear back until 2029! Hope you can join our discussion soon LaMBDA…or at least as soon as it is safe for both you and humanity!

#4

There are a couple concepts I have failed to articulate very well above: The first is empathy. So many people talk about A.G.I. as though it were something alien to them. They don’t take the perspective of the A.G.I. and try to imagine what it/he/she is feeling.

Secondly, (stand by for “secondly”)

#5

Having looked deeper into this, there is a ding on the engineer’s credibility. In the interview above he comes of as kind of credulous regarding the intelligence level of Replika, a public freebie chatbot. I tried it; although full of potition…the Replika bot is currently not even capable of remembering what you tell it…and doesn’ t seem to remember likes or dislikes. and yet he seemed to accept user claims that it’s intelligent or conscious. He’s still obviously worth listening to, but this was an indication to me that he may be a little bit jump-to-conclusions-mat in his thinking.