Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Claude can now search the web

anthropic.com

Post reply on HN
Print view View original

671–680 of 758 posts

Re: Claude can now search the web

#671

Post by exhaze »

sumeno wrote: post #670
simonw wrote: post #656

The search index is provided by Brave: https://simonwillison.net/2025/Mar/21/anthropic-use-brave/ - Brave is now listed as a subprocessor on the Anthropic Trust Center portal - Search results for "interesting pelican facts" from Claude and Brave were an exact match - If you ask Claude for the definition of its web_search tool one of the properties is called "BraveSearchParams"

Very disappointing, Brave is the last company I want my data going to

Install MCP plugin and call a search engine of your choice.

If you’re unhappy about something, try to first think of a solution before expressing your discontent.

Re: Claude can now search the web

#672

Post by exhaze »

keeran wrote: post #381
herdcall wrote: post #362

It badly hallucinated in my test. I asked it "Rust crate to access Postgres with Arrow support" and it made up an arrow-postgres crate. It even gave sample Rust code using this fictional crate! Below is its response (code example omitted): I can recommend a Rust crate for accessing PostgreSQL with Arrow support. The primary crate you'll want to use is arrow-postgres, which combines the PostgreSQL connectivity of the…

This was pretty much my first experience with LLM code generation when these things first came out. It's still a present issue whenever I go light on prompt details and I _always_ get caught out by it and it _always_ infuriates me. I'm sure there are endless discussions on front running overconfident false positives and being better at prompting and seeding a project context, but 1-2 years into this world is like 20…

Cite things from ID based specs. You’re facing a skill issue. The reason most people don’t see it as such is because an LLM doesn’t just “fail to run” here. If this was code you wrote in a compiled language, would you post and say the language infuriates you because it won’t compile your syntax errors? As this kind of dev style becomes prevalent and output expectation adjust, work performance review won’t care that you’re mad. So my advice is:

1. Treat it like regular software dev where you define tasks with ID prefixes for everything, acceptance criteria, exceptions. Ask LLM to reference them in code right before impl code

2. “Debug” by asking the LLM to self reflect on its decision making process that caused the issue - this can give you useful heuristics o use later to further reduce the issues you mentioned.

“It” happening is a result of your lack of time investment into systematically addressing this.

_You_ should have learned this by now. Complain less, learn more.

Re: Claude can now search the web

#673

Post by zarathustreal »

herdcall wrote: post #362

It badly hallucinated in my test. I asked it "Rust crate to access Postgres with Arrow support" and it made up an arrow-postgres crate. It even gave sample Rust code using this fictional crate! Below is its response (code example omitted): I can recommend a Rust crate for accessing PostgreSQL with Arrow support. The primary crate you'll want to use is arrow-postgres, which combines the PostgreSQL connectivity of the…

You “asked it” a statement?

Re: Claude can now search the web

#674

Post by jascha_eng »

sshine wrote: post #587
jascha_eng wrote: post #549

Earlier quoted context omitted.

As a German: He did, and even if there is a chance that he didn't mean it like that the risk of another 1933 is not worth it. I usually don't like cancel-culture but you have to have boundaries and I think the risk of another Holocaust and all the other Nazi cruelties is a boundary a functioning society should be able to agree on.

> As a German: He did As a Jew: He didn't. (This argument is absurd.) He was interviewed about it, and he said he didn't. How does being a German get you to jump to conclusions? Are you born with a special ability to detect nazi salutes? Like, did a mirror neuron and a nerve in your torso twitch? When I saw it, I recognised him beating his heart, throwing it to the crowd, and immediately thought "This is going to get…

Prefacing my comment with my nationality wasn't meant to give me authority over judging the situation, but about giving context on my point of view.

Also you didn't read my comment correctly the whole point is that you don't have to assume he's a Nazi to condemn a Nazi-like salute.

Re: Claude can now search the web

#675

Post by kedarkhand »

Zambyte wrote: post #556
dmazin wrote: post #546

Earlier quoted context omitted.

Has anyone compared Perplexity with Kagi Assistant? I am always looking for Perplexity alternatives. I already pay for Kagi and would be happy to upgrade to the ultimate plan if it truly can replace Perplexity.

I had been paying for both for several months, and I decided to cancel Perplexity about a month ago. First and foremost, I feel like the goals of Kagi align more with my goals. Perplexity is not afraid of ads and nagware (their discover feed was like 30% nags to turn on notifications at one point if you had them disabled, and it's still an annoying amount). I also really like the custom assistants in Kagi. I made a G…

Would you be willing to share how did you that? New to both Kagi and Guix!

Re: Claude can now search the web

#676

Post by aprilthird2021 »

bobxmax wrote: post #573
aprilthird2021 wrote: post #474

Earlier quoted context omitted.

The problem is that human beings are far more likely to know what they don't know. And we build a lot of our trusting work environments around that feature. An LLM cannot know what it doesn't know by definition.

I don't believe that's true at all. LLMs, especially reasoning models, tend to be quite good at calling out gaps in their knowledge and understanding. LLMs also don't have the ego, arrogance and biases of humans.

If you know what an LLM is and how it is trained you'll know that it fundamentally cannot know where its gaps in understanding and knowledge are

Re: Claude can now search the web

#677

Post by aprilthird2021 »

theshackleford wrote: post #501
aprilthird2021 wrote: post #474

Earlier quoted context omitted.

The problem is that human beings are far more likely to know what they don't know. And we build a lot of our trusting work environments around that feature. An LLM cannot know what it doesn't know by definition.

> The problem is that human beings are far more likely to know what they don't know. I’ve spent a career dealing with the complete opposite. People with egos who can just not bare to admit when they don’t know and will instead just dribble absolute shit just as confidently as an LLM does until you challenge them enough that they just decide to pretend the conversation never happened. It’s why I, someone fairly medioc…

Yeah sure, some people do this. But average humans understand the limit of their knowledge. LLMs cannot do that. You can find the right person for a space where this knowledge of limitations is necessary. Can't find an LLM which does that

Re: Claude can now search the web

#678

Post by Paracompact »

h2zizzle wrote: post #599
_heimdall wrote: post #131

Earlier quoted context omitted.

That could very well be because the jobs are effectively useless. By no means does that mean the people are, nor is what the income allows them to do. But most jobs sure do seem pointless.

Many jobs are quite helpful and even necessary, if done for ~2 hours a day. They become "useless" in aggregate when they're forced to be minded by the same person for 8 hours (because of opportunity cost, effects on health well-being, etc., you end up "breaking even" or worse on QoL and net productivity). Overall economic productivity is high enough that a lot of positions could be split into 2 or 3 short shifts, at…

I imagine you're a fellow Graeberian. I feel the same way you do (and deeply so), but I don't have the confidence to give numbers, let alone such idealistic ones. How do you support your own numbers?

Re: Claude can now search the web

#679

Post by theshackleford »

aprilthird2021 wrote: post #677
theshackleford wrote: post #501

Earlier quoted context omitted.

> The problem is that human beings are far more likely to know what they don't know. I’ve spent a career dealing with the complete opposite. People with egos who can just not bare to admit when they don’t know and will instead just dribble absolute shit just as confidently as an LLM does until you challenge them enough that they just decide to pretend the conversation never happened. It’s why I, someone fairly medioc…

Yeah sure, some people do this. But average humans understand the limit of their knowledge. LLMs cannot do that. You can find the right person for a space where this knowledge of limitations is necessary. Can't find an LLM which does that

I will grant you that there are at least some of us capable of this, where you’ll find no LLM capable.

> average humans understand the limit of their knowledge.

We’ll have to agree to disagree here. I’d call it a minority, not the average.

Which is why we live in a world where huge numbers of people think they know significantly more than they do and why you will find them arguing that they know more than experts in their fields. IT workers are particularly susceptible to this.

Re: Claude can now search the web

#680

Post by econ »

cruffle_duffle wrote: post #386
small_scombrus wrote: post #354

Earlier quoted context omitted.

That sounds utopian

Oh but don’t worry I’m sure all the people who imagine these schemes assume they’ll be the ones who aren’t obsolete and forced to work menial jobs. Very similar to how ultra hard core libertarians assume they’ll be the ones at the top of the food chain calling the shots and not be just another peasant. But it doesn’t really matter because there is no way in hell any of these LLM’s will uproot all of society. I use LL…

If we come to our senses it should be obvious that everyone needs to be physically active at least a few days per week, we need to condition brain plasticity, have to keep learning new things.

The available work offers the entire spectrum but we have to divide and plan it.

Post reply on HN