Live data from Hacker News

LaMDA: Google's New Conversation Technology

blog.google

61–70 of 157 posts

Re: LaMDA: Google's New Conversation Technology

#61
post #47
post #14

Earlier quoted context omitted.

IME, Assistant did do a whole lot more than Siri or Alexa in the past. It had a similar set of features, but with Assistant I could speak casually and with Alexa/Siri it was necessary to learn specific trigger phrases. More recently I find that Alexa has caught up. And maybe Assistant has gotten worse. Not sure about Siri.

Can assistant make calls to book your haircut with a real person yet? That was hugely exciting at a google io demo about 3-4 years ago

I used it to make reservations at an Indian restaurant a few days ago. I think it's still limited to the restaurant use case.

Re: LaMDA: Google's New Conversation Technology

#62

Disclosure: work at Google, opinions my own. I've had a chance to play with this and it's eerie how good it is. Most impressive to me, it was able to defend itself persuasively when confronted with evidence of how its previous answers were inconsistent (in the version I played with, it tends to be very coherent but frequently wrong).

Does it have to defend itself a lot?

I've noticed contradictory statements are kind of standard for GPT-3. I can well believe you could come up with a system that's better at defending itself against the charge of contradiction than it is at avoiding contradictions in the first place.

Re: LaMDA: Google's New Conversation Technology

#63

Disclosure: work at Google, opinions my own. I've had a chance to play with this and it's eerie how good it is. Most impressive to me, it was able to defend itself persuasively when confronted with evidence of how its previous answers were inconsistent (in the version I played with, it tends to be very coherent but frequently wrong).

Does it have to defend itself a lot? I've noticed contradictory statements are kind of standard for GPT-3. I can well believe you could come up with a system that's better at defending itself against the charge of contradiction than it is at avoiding contradictions in the first place.

In this case, I was pushing it quite a bit to see how it would respond. Most of the time, it responds quite confidently so it sounds right, even if it isn't, ie about the median of what you'd expect for a Hacker News comment.

Re: LaMDA: Google's New Conversation Technology

#64
post #31

Earlier quoted context omitted.

What would your response be?

"Hey, that's nice", "How's that going?" or something like it. Also, no response is also a response, you could just listen to them. No one's going to say "I just started taking guitar lessons" and remain silent for the rest of the evening; for sure they have much more to say about it.

Wait.

Why is it self-absorbed for someone to include themselves in the conversation, but it's not self-absorbed for you to come up to me and talk to me about your guitar playing?

You can come up to me and say, "Hey, guess what!? I just started playing guitar!" And I can't say, "No kidding? My mom just started playing guitar, too!" or, "No way! I just started drumming a few weeks ago. Maybe we can jam soon?"

How is my response "self-absorbed" but you starting a conversation about yourself isn't?

If you start the conversation, I'm just supposed to keep asking you questions about your experience until you're done talking about it?

I mean, if we're friends, shouldn't you care about what I think is relevant as much as I should care about what you think is relevant?

And anyway, if you _really_ want to talk about how it's going - can't you just take the conversation there?

I understand if you say something like, "Nice. I've always wanted to learn to play, and I never thought I could, and already I've learned so much, and I'm getting close to being able to play my favorite song."

And I say, "Oh, cool. Hopefully you can play my favorite song next. But it's really hard, you'll probably need a few years. You know, I used to be friends with Kurt Cobain's assistant, right? That's how I got into music in the first place....."

^ That seems self-absorbed. The first seems like genuine conversation.

Re: LaMDA: Google's New Conversation Technology

#65

Disclosure: work at Google, opinions my own. I've had a chance to play with this and it's eerie how good it is. Most impressive to me, it was able to defend itself persuasively when confronted with evidence of how its previous answers were inconsistent (in the version I played with, it tends to be very coherent but frequently wrong).

> Disclosure: work at Google, opinions my own.

Disclosure: don't care.

why do only google employees do this?

Re: LaMDA: Google's New Conversation Technology

#66

Disclosure: work at Google, opinions my own. I've had a chance to play with this and it's eerie how good it is. Most impressive to me, it was able to defend itself persuasively when confronted with evidence of how its previous answers were inconsistent (in the version I played with, it tends to be very coherent but frequently wrong).

> Disclosure: work at Google, opinions my own. Disclosure: don't care. why do only google employees do this?

It's not only Google employees.

And if people don't, others jump in to point out that they are "only saying good things because they work there".

Re: LaMDA: Google's New Conversation Technology

#67

Disclosure: work at Google, opinions my own. I've had a chance to play with this and it's eerie how good it is. Most impressive to me, it was able to defend itself persuasively when confronted with evidence of how its previous answers were inconsistent (in the version I played with, it tends to be very coherent but frequently wrong).

> Disclosure: work at Google, opinions my own. Disclosure: don't care. why do only google employees do this?

Disclosure: I work at Google too and my opinions are also my own.

We are told internally that when we discuss Google related matters online that we should be cognizant of the what/where/how/why of our communication. Disclosing is—for me—both a reminder for myself and an attempt at a hedge against accidentally representing Google.

Re: LaMDA: Google's New Conversation Technology

#68
post #66

Earlier quoted context omitted.

> Disclosure: work at Google, opinions my own. Disclosure: don't care. why do only google employees do this?

It's not only Google employees. And if people don't, others jump in to point out that they are "only saying good things because they work there".

they should add it to their profile.

Re: LaMDA: Google's New Conversation Technology

#69

Earlier quoted context omitted.

> Disclosure: work at Google, opinions my own. Disclosure: don't care. why do only google employees do this?

Disclosure: I work at Google too and my opinions are also my own. We are told internally that when we discuss Google related matters online that we should be cognizant of the what/where/how/why of our communication. Disclosing is—for me—both a reminder for myself and an attempt at a hedge against accidentally representing Google.

Sure, you should be cognizant of what you say online otherwise you'll get smacked with the cancel culture hammer. This is regardless of where you work.

Re: LaMDA: Google's New Conversation Technology

#70
post #66

Earlier quoted context omitted.

It's not only Google employees. And if people don't, others jump in to point out that they are "only saying good things because they work there".

they should add it to their profile.

They have. That's where the people who then gleefully point it out go to find it out.
Post reply on HN