Hacker News Reader

Comment by neosat | Hacker News Reader

The blog post is an example of why when you go through a number of edits and additions from various internal groups, marketing, legal, compliance etc. you end up with a message that's a lot of words but not a lot of information for any key audience.

I went through the article 3 times thinking I missed the link to try out Bard, get some sense of timeline, roll-out plan...nothing. Come on Google...

View on HN

It seems like a very reactionary move. ChatGPT went viral, and now they have to rush something to market to try and compete. So of course the product doesn't exist yet, but they've had enough time to put together a project plan and some mockups. It's not a great position to be in.

They've been doing internal testing for while, but I guess they never thought to release it publically until chat GPT became public.

There was even someone who thought it was sentient! https://www.cnn.com/2022/07/23/business/google-ai-engineer-f...

It reminds me of a blander Yann Lecun-like response where he responded on Twitter to a bunch of people saying Facebook ai could have done this and that, without acknowledging they were beat to the (shipping) punch and are now just reacting. Building a roadmap on what is in the zeitgeist and trying to pretend you’re winning a fight you don’t realize you already lost (at least the first battle) is not a way to capture mindshare imho

What are you referring to? My overall impression of Yann Lecun is he feels quite hamstrung by all the extreme caution at Facebook (which obviously exists at Google too) of releasing anything that could produce content that is in any way objectionable to one of a million different subgroups.

I like Yann and think he’s brilliant and also generally appreciate how he engages on Twitter. However, he spent a lot of time recently downplaying openai and I don’t see why that was needed other then to try to say FAIR is important, see one of the most recent of these related tweets https://twitter.com/ylecun/status/1620533783702433792

It does seem slightly amateurish. If there was more self-confidence like Apple, there wouldn't be a need to announce anything until it's ready.

EDIT: Especially if it was already available internally for several months as some others say, waiting another month or two until it's ready to go doesn't seem like a big burden.

Google was in this position in relation to Apple and the iPhone, and that worked out all right for them. First-mover advantage isn't the be-all, end-all.

I think the real fear is that there has been a brain drain of high-level AI talent to well-funded startups because those startups can move faster without having the extreme fear/caution of a large company like Google/Facebook. E.g. my guess is that OpenAI doesn't have an "Ethical AI" team who sees their role as primarily to put up roadblocks to releasing.

Hence not surprised by the big investment in Anthropic by Google.

There was a different dynamic there. Apple was clearly focussed on the higher end of the market (price point) and unwilling to compromise on profit margins, while willing to sacrifice pure market share. Google took advantage of that to win the distribution space focussing on lower price point phones on Android (and then expanding to higher price point devices).

With ChatGPT the threat is different and both are going after expanding usage.

I don't think so. Google has had its own internal LLMs for quite some time now. Some of them are even more powerful than GPT 3.5. The only question has been: "Why release it to the public?"

Exactly. I played with "Meena" when I was an employee there and it was ... disturbingly human-like. I can understand why Google didn't want to release it. And even more so after the Blake Lemoine incident.

Google is already accused of operating a panopticon, the last thing they want to be accused of is running SkyNet.

People who think Google has fallen behind here are sorely mistaken. They just don't/haven't-had a way to make money off of it and are likely worried about reputational fallout.

I want Google's lunch to get eaten as much as the next guy, but I don't think it will be on this front.

It is a reactionary move, but Bard existed for months on the intranet, but not as a public offering. So they have the technology but not the product.

Presumably this was the infamous tech that got that one worker fired because he thought it was sentient and wanted to blow the whistle on unethical AI treatment right? This story: https://www.cnn.com/2022/07/23/business/google-ai-engineer-f...

It would be funny if Google hastily rehires him as a marketing evangelist to show people their AI is so good you think it's alive.

While I don't think LaMDA is sentient (merely based on what I understand about its architecture), what would it look like if some mega corp had a conscious AI on its hands, and an insider was trying to blow the whistle?

Wouldn't it look exactly like this?

In one sense, yes, but to me that felt more like a mental illness story than anything else, which would make it much less funny (apologies to the person involved if it was not).

It is a bit surprising that Google got scooped by OpenAI, but the product surely has existed before ChatGPT went public. LaMDA was the model behind the 'Our AI has become sentient' controversy last year, if you remember that [0].

[0] https://www.washingtonpost.com/technology/2022/06/11/google-...

It's not rushed to market. It already existed for years.

I've tried to say this before elsewhere, but Google has had something internally that's competitive with ChatGPT already for years, under various names. They were just naturally reticent about letting it loose on the world. Esp after the Blake Lemoine incident.

Exactly. With their partner Anthropic's bot "Claude" they have something on the same level as ChatGTP already though.

You can test Claude on ios with the app "poe" (the other two bots in the app are some sort of chatGTP and OpenAI based).

i don't think anybody knows how to monetize it.

really it's kinda a public service, maybe these models should be run that way?

Monetization seems pretty straight forward.

User: What is the most durable shoe?

AI: Some of the most durable shoes are ... Here are some affiliate links where you can buy these durable shoes...

Just picture monetizing it via embedding slant onto the answers it provides like how Alexa suggest purchases but in a way that users can barely just notice...

Or purchasing "market share space" on it in the same way that companies buy shelf space on supermarkets to place their products in...

as someone who did nlp research at Google, Google had a ChatGPT level chatbot 3 years ago and did nothing with it

This sounds a bit like those stories about Kodak having a digital camera before anybody else did, and they just fearfully watched the thing from a distance doing nuthing until someone else brought it to market and ended the party for them.

May be an entirely different thing this time of course. But still...

Well they had a giant conversational size bigger than GPT3. It was not as safe or good at following instructions as ChatGPT. But it was as flowing as it is. So maybe it's an exaggeration (this word has two gs in english!?) to say that it's as good, but it was also extremely strong and impressive

Got you: https://labs.kagi.com/ai/sum?url=https://blog.google/technol...

  Sundar Pichai, CEO of Google and Alphabet, has announced the release of Bard, an experimental conversational AI service powered by Google's Language Model for Dialogue Applications (LaMDA). Bard seeks to combine the breadth of the world's knowledge with the power, intelligence and creativity of Google's large language models. It draws on information from the web to provide fresh, high-quality responses. Bard is initially being released with a lightweight model version of LaMDA, which requires significantly less computing power and will allow for more feedback. Google is also working to bring its latest AI advancements into its products, starting with Search, and will soon be onboarding individual developers, creators and enterprises to try its Generative Language API. Google is committed to developing AI responsibly and will continue to be bold with innovation and responsible in its approach.

Wow, thank you for that link. Now I have a new bookmarklet...

javascript:location.href='https://labs.kagi.com/ai/sum?url='+encodeURIComponent(locati...

Formatting is a little messed up, but great idea here.

Thanks for the bookmarklet!

They probably launched this for free as marketing to increase visibility into their startup! Enjoy it while it lasts before the service is paywalled

Note I've been using Kagi happily for several months and it has successfully replaced Google Search for me. Highly recommended.

Exactly, he announced that they are doing a thing. With a very ambiguous screenshot of the thing and zero evidence of just how good the thing is. And no link to the actual thing.

So, the thing doesn't exist yet in a form that is concrete or demoable. And it definitely isn't ready for users. Which is the same thing really. Also there's no timeline of the thing actually getting there either. So, there's nothing here really.

Why is Sundar Pichai still in charge of this company? Months of excitement around chat gpt and then the best he came up with is this?! This reads to me like "The dog ate my homework, sorry. I have nothing of substance to announce today. Or tomorrow. Or any time soon.".

Also, Bard. Really?! Cringeworthy doesn't begin to describe how bad that is as a brand name. It' sounds like Bad spelled wrong.

Having worked in big tech for a while, my written-by-a-committee alert went off quickly for this. Founders are the only ones who seem to get away without doing this when running large companies. Personality and substance are very important for communication and committees (and the folks who want to edit as part of one) never get that right.

There is nothing concrete available today because Google is scrambling to get ahead of a major announcement from Microsoft on integrating GPT into Bing. Internal sources from MS are saying it might be the first we see of GPT-4.

Why do you say google is scrambling when they have been working on AI stuff just as impressive as OpenAI, which you could spin differently to say Microsoft “scrambled” to make a partnership with and are now attempting to cram GPT into their tech stack.

FWIW I don’t think either company are really scrambling except in the performative sense of making announcements to appease the market.

Microsoft's partnership with OpenAI is a years-long affair, not something they're just starting on now. Google should have seen this coming and been able to beat them to release, but they seem to have been caught off guard with some great research projects but no product.

Both companies have been doing AI work for a long time. Do you really think google is behind on AI here just because they haven’t released any products? What do they stand to gain by pushing out half-baked toys? They have nothing to prove.

Google isn't Apple. Google's strategy for the last 10 years at least has been to throw out half baked products and see what takes off. If they had anything usable, we would know by now.

And, on the contrary, Google has everything to prove. ChatGPT exploited a years-long dissatisfaction with Google search and has millions of people using it in lieu of Google's primary product. This is the most existential threat that Google has faced since its birth, and they are not handling it well.

You’re living in a fantasy land. If I had a major criticism of Google in the past decade it’s that they don’t release early or often enough.

If you think people are dissatisfied with google search then you’re missing the point that people don’t think about google search at all, they just reflexively use it all day. I don’t know anyone who uses chatgpt with such frequency or in a way that is so central to their daily life, and I have a much more tech-savvy circle of friends than most people.

Because while Google has been researching, OpenAI has been productizing. We will have to judge it once we get access to Bard, but it’s possible Google is years behind on creating a chat model that won’t spew toxic waste, as GPT-3 did when it was first released a few years ago.

Not spewing toxic waste is actually a big part of what Google has spent years working on here. The fact that someone else released the first such product allows them to do it too without solving the problem with less media flak.

Productionizing this stuff is where Google gets the most advantage because they have the hardware and software efficiencies that comes form years of experience training and running inference on the most massive AI workloads for many many years in their data centers

I don't think this is going to be what damages Google, much more optimistic about antitrust stuff.

Yeah I keep thinking about how different the reception to ChatGPT would have been if it were released by Google. People would be way more focused on how you can make it spout total nonsense with supreme confidence. (Which imo is the major flaw in all these ML models.)

I work at google and have used their AI tools and they’re just as impressive as the OpenAI ones (which I use personally, outside of work). Both have been developing products, just google hasn’t released any of them.

Have you done adversarial testing? How easy is it to jailbreak Bard vs. ChatGPT? Can Bard produce code?

Obviously I am unable to discuss the details, as much as I might like to. Sorry.

That’s just it though. GPT-3 has been publicly accessible for years. ChatGPT has been public and hitting scale for months. OpenAI has collected mountains of AI chat training data and has been iterating in public. Google has nothing available to the public and you are under NDA. And you think we should be optimistic about Google’s progress? Will this new search feature last longer than Stadia?

Either way, chat interfaces in both search engines should be available to some of the public within the next few weeks. Theory is about to smack into reality at scale.

It's corporate-speak for 'we're not letting the threat going unanswered' without actually delivering anything yet. It's mostly a placeholder for shareholders and especially the media to move the story away from 'Google's in trouble'

Maybe the Bard announcement was written by Bard?

Out of curiosity, I copied all the text from the website (so it got header, footer, unrelated text) into ChatGPT and asked it to summarize:

This is an article written by the CEO of Google and Alphabet, Sundar Pichai, about the company's journey with Artificial Intelligence (AI). Pichai discusses how the company has been working on AI for the past six years and how they have been advancing the state of the art in the field. The article mentions the release of an experimental conversational AI service called Bard, which seeks to combine the breadth of the world's knowledge with the power and intelligence of Google's large language models. The CEO also talks about how the company is working on bringing the benefits of AI into its everyday products, starting with Search, and how AI can deepen people's understanding of information and turn it into useful knowledge more efficiently.

There is an event on Wednesday which will likely provide more information.

There is no information to give. This "announcement" is a reaction to all the stories about how ChatGPT is going to eat Google's lunch. The company needed to do something to defend against that in the market - for the sake of its shareholders and its advertising revenue.

That said, here is the chatgpt summary I generated:

Google CEO Sundar Pichai announces the release of their conversational AI service "Bard," which combines the world's knowledge with the company's large language models. Bard seeks to provide fresh, high-quality responses to questions and allow users to explore new information. The release is part of Google's effort to bring the benefits of AI into everyday products and deepen people's understanding of information.

This kind of no information exec-speak is very painful to read. I had to pass it through ChatGPT to summarize. THe whole article can be summarized in 1 sentence without any loss of information: "Google will name it's chatbot Bard and it be based on Lambda model, initially small size". No link, no other details, basically a lot of words to say nothing at all.

If you have to ask, you're not invited.

We know how well that worked out for Google+

It worked for gmail. Maybe some people are too young to remember it was invite-only for years.

That's because there was no competition and they were the first movers.

For free email? That is patently untrue.

No one came close to gmail's 1GB free inbox at the time. Which kept increasing and showed you at the bottom of the inbox. Plus it had amazing spam detection, and was stupid fast, and got better and better very quickly.

I remember having a gmail email was almost a litmus test to whether or not you were paying attention to what was going on in tech. Not having a gmail email indicated you might not be keeping up with the times.

Agreed on all your points! But I think offering a better service in an established sector is the opposite of "no competition" and being a "first mover".

Like how we would say a professional sports player playing against an amateur has no competition. Technically, yes, but effectively...

The context of the discussion is whether or not an "invite only" tactic works for building hype and launching a successful product. In that sense, I would attribute Gmail's success not to the "invite only" tactic, but to the fact that it was so superior that it had no competition.

Hotmail preceded gmail by most of a decade.

I had a Hotmail account for about a year in 2000 or so. It had a 10 MB inbox and worked quite well. I got a bit of spam, but it was obvious and manageable.

My girlfriend got a GMail invite from a friend and sent me an invite. It was huge. A 1 GB inbox, built-in Google search of your email, and incredible spam filtering.

They marketed it something like "With GMail, you don't have to delete your emails. Just search . . .". Really killer features compared to their competitors.

Later on, they added automatic email thread grouping which made using Outlook for work a chore until it got a similar feature.

Not with a 1GB (!!!!!!) inbox.

Coincidentally also why ChatGPT won.

Gmail was objectively far, far superior than all the other competing email offerings, so people wanted to be invited.

I remember asking for an invite on what was probably a phpbb.

Apparently the "trusted testers" are basically all internal employees

>CEO Sundar Pichai told employees Monday the company is going to need all hands on deck to test Bard, its new ChatGPT rival. He also said Google will soon be enlisting help from partners to test an application programming interface, or API, that would let others access the same underlying technology.

https://www.cnbc.com/2023/02/06/google-ceo-tells-employees-i...

It means they are not a “trusted tester.”

Or they asked their new AI to write the post and they got a result that sounds amazing but is still total fluff

> that's a lot of words but not a lot of information for any key audience

So they used ChatGPT to generate it?

> we’re taking another step forward by opening it up to trusted testers ahead of making it more widely available to the public in the coming weeks.

this is in like the 5th paragraph, right under the 'Introducing Bard' title.

Have you ever tried utilize one of their APIs..?

It's an atrocious experience. They're leaving so much money on the table.

Also good luck signing up and making your first API call.

OpenAI made that very easy, Google no doubt will make it a total pain.

Still can't find the link.

"making it more widely available to the public in the coming weeks."

It was written by ChatGPT

Or maybe they had ChatGPT write their blog post. ha ha ha...

Reply on news.ycombinator.com