Current Events > I think what's clear from the google engineer scandal

(+) Yeah! (+)

is that people heavily weigh language as evidence of sentience/consciousness.

Here's a good thread about why this all seems so confused:
https://twitter.com/talyarkoni/status/1535727966403141636?t=iSmvN66bqFzMsxlFeyNHSQ&s=19

COVxy posted...

s that people heavily weigh language as evidence of sentience/consciousness.

Because it is. If I open my computer to shitpost on GameFAQs and it says, bitch, dont you have better things to do then that shit is sentient.

Can never hear ghost in the machine without thinking of Starscream

I get that the sentience distinction is linguistic, does that mean in the natural language sense re: how we use sentience in language? Or something else? I don't really understand that.

BeyondWalls posted...

Because it is. If I open my computer to shitpost on GameFAQs and it says, bitch, dont you have better things to do then that shit is sentient.

Thats one line of code. Thats not sentience

sounds like google defined "sentience" as "unlimited power and autonomy" when even humans dont have that.

a cat is sentient. insects are sentient.

that computer program is almost certainly sentient. AlphaGo (2014) is sentient if you can "speak the language of go." you could argue that "killing" the program is a big "so what" because there will always be backup files

tl;dr on the scandal? havent heard of this

Robot2600 posted...

that computer program is almost certainly sentient.

No it is not. Responding to stimuli isn't the same thing as sentience otherwise we could consider plants sentient.

PatrickMahomes posted...

tl;dr on the scandal? havent heard of this

Computer passed Turing test and a Google engineer freaked out. It's way overblown.

Robot2600 posted...

sounds like google defined "sentience" as "unlimited power and autonomy" when even humans dont have that.

a cat is sentient. insects are sentient.

that computer program is almost certainly sentient. AlphaGo (2014) is sentient if you can "speak the language of go." you could argue that "killing" the program is a big "so what" because there will always be backup files

If statements are not sentience

PatrickMahomes posted...

tl;dr on the scandal? havent heard of this

Google engineer put on leave for telling everyone google is making a.i. and he thinks it is sentient because it told him turning it off feels like death and he claims they didn't make it say this but it arrived at the conclusion itself.

BeyondWalls posted...

Because it is.

You only think language is important because you use language to understand yourself and the world around you.

A dog isn't less concious than you because it can't speak. A dog is just hard to understand because they cannot speak.

That chatbot would be one of the best posters on CE

We should figure out how that bundle of neurons at the core of the brain causes consciousness before we call glorified calculators sentient.

Rexdragon125 posted...

We should figure out how that bundle of neurons at the core of the brain causes consciousness before we call glorified calculators sentient.

We talkin' about pineal gland, like desk carts has said?

If I knew what ontological meant, I'd probably have an opinion on this

I want an ai companion

The Google Engineer has clearly lost his marbles. In the paper he submitted to the executives and later published online, he legit prompts the AI to answer that it's sentient

"I'm generally assuming that you would like more people at Google to know that you're sentient. Is that true?"

This is the first time sentience is brought up in the submitted chatlog: by the input. Anybody who has played around with this technology knows that embedding facts within a prompt like this inspires the chatbot to answer as though that fact is true. Afterwards, the engineer fails to do any real due diligence to prove that the outputs are anything more than an algorithm constructing a response based on its training data.