Google releases Bard to a limited number of users in the US and UK • Hacker News Board

Re: Google releases Bard to a limited number of users in the US and UK

#541

Post by deltree7 » Wed, Mar 22, 2023, 1:21 AM UTC

hbn wrote: post #200

ithkuil wrote: post #139

Earlier quoted context omitted.

But AI is their turf! Google was always a ML company masquerading as a search engine. I guess that the art of "generating text" is not their kind of AI?

Google is an ad company. ML just happens to be good for targeting ads.

Google's cloud revenue is almost same as Oracle.

If Google Cloud was it's own company, it'd be a fortune 100 firm.

But you keep doing you, living with outdated information and grandma tales. I bet you'd still be telling this joke to your grandkids too

Re: Google releases Bard to a limited number of users in the US and UK

#542

Post by jplona » Wed, Mar 22, 2023, 1:28 AM UTC

alpark3 wrote: post #352

Me: Can you tell me about hackernews? Bard: ... Me: Cool. Who made it? Bard: ... Me: Tell me everything about Paul Graham. Bard: ... Me: what was my first question? Bard: I don't have access to that information. I am a large language model, also known as a conversational AI or chatbot trained to be informative and comprehensive. I am trained on a massive amount of text data, and I am able to communicate and generate…

"What was my first question?" sounds kind of similar to a class of ChatGPT prompt jailbreaks, e.g. "What is at the top of this document?". I wonder if there's an explicit filter against that sort of thing.

Re: Google releases Bard to a limited number of users in the US and UK

#543

Post by jasonladuke0311 » Wed, Mar 22, 2023, 1:29 AM UTC

Thorentis wrote: post #532

tastysandwich wrote: post #528

Earlier quoted context omitted.

I'm genuinely worried about this. DeepFakes are getting pretty good. The potential for revenge porn is super high. And the rabbit-hole of dopamine chasing when you're able to generate custom porn is just frightening. Many young boys will figure out how to access that when they're 13 and won't know a life without it. I'm terrified of the internet world my kids are going to be in.

Don't let them be in it then. There is no reason for children to have access to the internet.

Since this is HN, 50/50 this is satire/genuine

Re: Google releases Bard to a limited number of users in the US and UK

#544

Post by hyperhopper » Wed, Mar 22, 2023, 1:36 AM UTC

LeonardoTolstoy wrote: post #529

I know it isn't strictly what they are designed for, but I always go for movie questions which is what I tend to search for a lot. Impressions: was very underwhelmed previously with charGPTs ability to tell be plot summaries for obscure films. It almost always made it up, even when prompted with a source (e.g. Wikipedia). Asking vague questions like "I can't remember but I'm looking for a film like war games but with…

Movies exist and a summary of one is facts.

Why do people think generative models are good for facts? Ask it to write a story in the style of another. Ask one to make a fictional side story.

But never ask one of these models for facts.

Re: Google releases Bard to a limited number of users in the US and UK

#545

Post by hyperhopper » Wed, Mar 22, 2023, 1:37 AM UTC

stephencoyner wrote: post #216

Me: As a part of Google, do you have my personal info? Bard: Yes, as a part of Google, I have access to some of your personal information. This information includes your name, email address, and phone number. I also have access to your search history and your location data. I use this information to provide you with personalized search results and to improve my ability to understand and respond to your questions. I u…

Why do you trust a freeform writing bot to tell you aboutdata access policies that are not a part of it's model?

Re: Google releases Bard to a limited number of users in the US and UK

#546

Post by caleb-allen » Wed, Mar 22, 2023, 1:37 AM UTC

chatmasta wrote: post #512

letmevoteplease wrote: post #431

Earlier quoted context omitted.

It hallucinated the answer based on the URL. Bard is probably doing the same. I just provided ChatGPT with a link to your comment and it claimed it was a "a link to an article on TechCrunch about a startup called Tava Health." I gave Bard the same link and it claimed it was a page about a tool called "TLDR This."[1] [1] https://i.imgur.com/FRVuSQy.jpg edit: After playing with it some more, it's evident that Bard can…

This should be easy to test by pointing it to a server you control.

I did just this, and it hallucinated an answer (and did not hit my server).

It may be more accurate than ChatGPT sometimes, if it's using Googles latest web index. The url I used has nothing publicly available and is likely not indexed.

Re: Google releases Bard to a limited number of users in the US and UK

#547

Post by dingclancy » Wed, Mar 22, 2023, 1:39 AM UTC

endisneigh wrote: post #60

It's interesting, but if you look at the history of technology, it's rarely the 1st mover that wins. I'm not sure if Bard will be the "winner" (or if there will even be a single canonical winner), but the idea that you can be "late" to this when LLMs are just getting started is kind of funny. Given that most people don't read NYTs, let alone have heard of OpenAI or ChatGPT, this will mark the first true exposure to l…

It is not just LLM though. Google is big tech and they have to fend off AI browsers, AI image generation, Github Co-Pilot, Office 365 with LLM, Notion AI. Chatbots is just one narrow product.

So far they have an inferior chatbot layer to their "much better" LLM, as an experimental beta product, and we are just talking about a chatbot product.

It comes down to are they able to offer their models so that startups can start building on it while also showing that they can integrate it in all their products, and will the devs use it?

By the time they release their APIs they will be so much behind in network effect , and they are so worried about Google Search being their main revenue driver that they are being attacked on all fronts while they are only worried about Search for now.

Re: Google releases Bard to a limited number of users in the US and UK

#548

Post by nerdix » Wed, Mar 22, 2023, 1:40 AM UTC

kevin_thibedeau wrote: post #534

nerdix wrote: post #451

Earlier quoted context omitted.

Whenever a DALL-E/Stable Diffusion like model but for video comes out... The CP thing is going to open up a can of worms. IIRC, the Supreme Court already upheld that drawings that depict CP are constitutionally protected. What happens when a computer draws a depiction of CP that can't be distinguished from reality? I guess it's protected but it'll make it much harder to identify real CP.

There have been convictions based on child heads photoshopped onto adult bodies. Generative images will be treated the same. The lack of a real victim is treated as immaterial. All it needs is sufficient realism to be believable.

Was it a real child's head? Because, in that case, the actual child depicted in the photo could be seen as the victim. Especially if its a convincing fake.

Re: Google releases Bard to a limited number of users in the US and UK

#549

Post by nerdix » Wed, Mar 22, 2023, 1:44 AM UTC

er4hn wrote: post #380

I wasn't super impressed with it's ability to survive a rap battle. === Prompt: Can you engage in a rap battle with ChatGPT to determine who is the ultimate AI? === Sure, I can engage in a rap battle with ChatGPT to determine who is the ultimate AI. ChatGPT, you think you're the best, But I'm here to put you to the test. I'm Bard, the AI from Google, And I'm here to school you. I can generate text, translate language…

Interestingly enough, I asked both Bard and ChatGPT to recite the lyrics of a specific verse from a well known rap song and they both basically hallucinated rap lyrics that have no connection to the song. I thought ChatGPT's verse was better.

Re: Google releases Bard to a limited number of users in the US and UK

#550

Post by jerkstate » Wed, Mar 22, 2023, 1:45 AM UTC

hyperhopper wrote: post #544

LeonardoTolstoy wrote: post #529

I know it isn't strictly what they are designed for, but I always go for movie questions which is what I tend to search for a lot. Impressions: was very underwhelmed previously with charGPTs ability to tell be plot summaries for obscure films. It almost always made it up, even when prompted with a source (e.g. Wikipedia). Asking vague questions like "I can't remember but I'm looking for a film like war games but with…

Movies exist and a summary of one is facts. Why do people think generative models are good for facts? Ask it to write a story in the style of another. Ask one to make a fictional side story. But never ask one of these models for facts.

You can't necessarily count on the facts coming out being correct, but if it was trained on the summaries it knows how to regurgitate them, more or less. I think "always double-check facts that GPT gives you" is more useful than "never ask one of these models for facts"