Hacker News Reader

Comment by sssilver | Hacker News Reader

If anyone's wondering how this works, it looks like[1] Gemini + guardrails:

> Google is proud to be named a technology partner in this vital initiative, leveraging Gemini to help more than 100 million people access critical public resources with greater speed and ease

[1] https://blog.google/company-news/outreach-and-initiatives/pu...

View on HN

Elon has said that it's Grok. Take that for what you will.

https://x.com/elonmusk/status/2105060152919662622

Do you not think we can take that at face value?

Why not?

He constantly spreads lies.

I think that's the joke.

Does he?

I wouldn't qualify bad predictions as lies. Is there really a "constant" set of lies outside of that?

Yes. He's quite fond of debunked election fraud claims, for example: https://www.reuters.com/world/us/elon-musk-delivers-cash-con...

He also continuously claimed DOGE actually saved taxpayers money when it objectively did not.

Elon did some great things, like bringing EVs back into public consciousness, and SpaceX really did push the aerospace industry back into research/discovery mode. However, his behavior on social media and in the political sphere has been unfortunately quite, and I emphasize "quite" here, horrible.

He literally hard coded his chatbot to spread lies about white genocide in south africa, that's off the top of my head.

They're lies when he makes these "bad predictions" knowing they won't come true. He has constantly committed fraud in order to get the stock price of his companies up.

He's a compulsive liar.

Musk has made dozens of verifiable lies publicly. Why would you take him at face value?

Well, because there's a chatbot powering america.gov, and there's only a few choices for models/companies--of which grok is one. And here Musk is claiming grok was chosen, while nobody else seems to be making an opposing claim.

> (…) nobody else seems to be making an opposing claim.

Did you skip the first comment of this thread?

This makes sense. I'm assuming that, for the AI search assistant, Google has worked plenty to refine data lookup and retrieval—they definitely have some internal tech that competitors don't.

It’s wrong

>google's secret sauce

https://en.wikipedia.org/wiki/Tf%E2%80%93idf

Take it from me, this'll be all the rage next year.

"Save token spend by optimizing your agents' grepping spree!"

> they definitely have some internal tech that competitors don't.

So this is why Google is considered so far ahead of its competitors in the AI race!! (lol)

Ah, I wondered. It stubbornly refused to admit what it really was, kept saying it was built by some never-heard-of group at the White House lol

This must be why I can’t provision compute in us-central1

So it's literally Google search for the government? The unstoppable force has met the unmovable object!

> Your personal information isn’t collected or stored, and your conversation disappears when you leave.

I am left to wonder how this is technically accomplished. Is inference running in the browser? Is the model downloaded to the user’s machine?

It most likely means that Google doesn't persist conversations.

I don't think an HTTP header would normally be considered "personal information", so I don't see why you imagine there would be a technical limitation here.

Personal information is not clearly defined so it can mean anything. My guess is the search request is not directly linked to a specific user but a session to which that session is linked to a user so there isn't an identifiable person making a search without some backend data matching.

Taking a quick look at the network traffic, the site does a POST to https://america.gov/api/chat with a JSON payload when submitting a query:

  {"language":"en","id":"VSvMeyCMKTX05oLW","messages":[{"parts":[{"type":"text","text":"how do I register my new business?"}],"id":"oHub5UDJ3iYARhao","role":"user"}],"trigger":"submit-message"}

So the user's query is hitting a server so to answer my own question, the answer is No to local inference / models.

Edit: PDF attachments and voice input are processed locally according to the site's privacy policy[1]. But they still use 3rd party 'AI Providers' that abide by a ZDR.

[1] https://america.gov/privacy-policy

ah, that explains a lot, I see the attitude resemblence

Reply on news.ycombinator.com