YHK - yet another hacker news clone

Amazon blocks Meta’s new Muse AI agent from shopping on amazon.com

55 points simianwords 1 hour ago 29 comments

majorchord 34 minutes ago | parent

Why is it even Meta's job to announce that an AI agent is being used? Is there a law that actually forces them to make that known?

Otherwise people could just use local agents on their own machine controlling their existing browser/logins as they likely already do, without issues.

JumpCrisscross 19 minutes ago | parent

I guess it’s more that Amazon’s server queries for an answer and intentionally obfuscating a bad one comes across as lying.

esafak 14 minutes ago | parent

What do you mean? Declaring the agent is transparency. You should be congratulating Meta for their rare good deed.

geodel 34 minutes ago | parent

Thank you Amazon for standing up for everyman Rufus. We don't need AI agent from evil ad giants.

majorchord 31 minutes ago | parent

How do you know they are standing up for users? From where I'm sitting it looks like they just are trying to protect Alexa Shopping's revenue and/or monopoly.

suprfsat 28 minutes ago | parent

Amazon: famously not evil ad giant.

kqgnkqgn 26 minutes ago | parent

I assume it was sarcasm, Rufus is an Amazon AI product.

a4isms 18 minutes ago | parent

And there go my Bill & Ted's Excellent Adventure memories. George Carlin's Rufus was a "good 'un."

digdugdirk 24 minutes ago | parent

Ha. Amazon's in-app AI agent is called Rufus. It's incredibly annoying, and they love directing you to an 'Ask Rufus' page where you watch an AI agent not-quite-finish repeating your search terms back at you before you realize you're not actually on your search results page, and then spend some additional time trying to find the link to take you to said search results page.

I assume somewhere there's a team hitting their AI query KPI's due to this, but I really wish they'd shut off a few server racks and save everyone the annoyance instead.

simianwords 31 minutes ago | parent

There's a lot to discuss here - agentic commerce, IMO, is where things are heading. I'm completely convinced that people _want_ to shop using agents. But I want to understand the dynamics at play here: will underlying inventory providers like Amazon give up control and allow agents to use their website?

I find this question very interesting and I want to predict where things will head. On one hand Amazon has a lot to lose by giving up control on how their website is used. Ideally they'd like to control the full end to end experience. What happens to the work they put in their UI? All those endless A/B experiments to make the buttons look perfect?

But on the other hand - this is where things are heading. They can't stop agents from using their website. Not for long. And they have a lot to gain because overall purchases will increase since. They can accept it and create a plugin and try to offer as much "branding" as possible inside it. Or they can completely go against agents and block it.

Then you have to wonder, what really is the moat here? Amazon's logistics and its relations with partners and customers (return policy etc) and the trust it brings. The UI part dissolves and is no longer important.

Maybe they want only an allowed list of agents access to Amazon? Maybe that's where things are heading? They sign deals with maybe just OpenAI (for example)? Its not like Amazon makes their own agent + model. I would hate it things fragment this much.

majorchord 28 minutes ago | parent

Anything is possible I suppose... I don't see a path that leads to Amazon actually further controlling the shopping experience inside a browser (where a local agent might have control that they can't reliably detect), short of forcing users into some kind of hardware-based attestation solution or something, but even that is not foolproof.

dylan604 16 minutes ago | parent

> On one hand Amazon has a lot to lose by giving up control

The everyday person is told to get over it and accept that AI is here so adapt or die. Companies start reducing headcounts to be replaced with AI. So now, a company is having to accept that people are using AI in ways they don't like so they should adapt or die as well too, right? I have no tears to shed if Amazon loses control. Also, it'll be one billionaire battling another billionaire at that point, and I'm kind of okay with that chaos. My world is no longer in my control so fuck 'em is my stance

simianwords 15 minutes ago | parent

OK according to game theory, if this is correct, why even put up a fight? You have nothing to gain by fighting. Just accept it and play along. Make your damn MCP. You have 100,000 engineers. Why can't connect Amazon to ChatGPT today? I don't know.

edmundsauto 10 minutes ago | parent

It’s an optimistic view on AI - brings power to the people by mediating all the garbage.

I’ve already used it to handle customer support request calls. I see 30+ minute hold+conversation times and generally the issue gets resolved. That is hugely empowering.

edmundsauto 11 minutes ago | parent

In theory, legislation could help. If there were a pro-consumer movement in government.

Something like declaring agents as representatives of consumers and blocking them violates ADA or just shouldn’t be allowed unless they are abusive.

Lacking that, it seems it’ll be an arms race between offense and defense. At some point, computer control probably wins (running my browser, on my computer, is pretty difficult to detect with reasonable f1 score)

Jcampuzano2 28 minutes ago | parent

Agents don't click through or interact with ads or get influenced by things like lightning deals, snap promotions, "others who bought X bought Y" and such the same way a human does. This is likely a large portion of the reasoning.

This is just going to get worse before it gets better. More large sites are likely going to block or attempt to block agents because they're so reliant on ads for revenue or exploiting human psychology.

simianwords 22 minutes ago | parent

Its a very good point. There are lots of orgs in company's whose whole job is to push shit from all possible points in the website - its probably called "real estate" internally.

These orgs have a lot to lose. But the company? Not so much, they will make it up and more - agents are great to use.

And yes ads are hard to place because agents let you bypass all this and give you a nice filtered view of exactly what you need.

Where do you think the world will head? Maybe Amazon will sell their inventory for agents at a hefty API price?

svachalek 6 minutes ago | parent

Yeah subscriptions are another big one. So many profits today are generated by bleeding off customers that just can't deal with the hoops they need to jump through to get unsubscribed. Tasking that to an agent would destroy the whole model.

United857 27 minutes ago | parent

There's a difference between automated industrial scale scraping and well-behaved agents acting on the behalf of individuals. Right now there isn't a robust, standard way to distinguish between them, so sites just block known datacenter IPs and throw out the baby with the bathwater. That's a main advantage of running your own local claw setup using your residential connection -- difficult/impossible to block.

That said, a site blocking all agents would be like blocking all search engines, something that probably hurts more long term than it helps as agentic interactions become "the norm".

dgellow 2 minutes ago | parent

> Right now there isn't a robust, standard way to distinguish between them

There is, it’s called an API

stephen_cagle 1 minute ago | parent

It's worse than that, they will block all agents and specifically make exceptions for maybe 2 search engines per country. Thereby strengthening existing players and weakening all incumbants.

Cloudlare for instance does allow indexing by Google and Bing I think by default, but does challenge other bots. Double check me I am speaking from memory.

stbenjam 25 minutes ago | parent

It is a money grab from Amazon I'm sure. They want to make a deal with meta to get paid for access. Maybe also worried about liability I imagine on accidental unauthorized purchases?

shagie 15 minutes ago | parent

Returns cost money and if an agent generates more returns and hitting the margins or seller retention then that's something that's likely going to get increased scrutiny.

The generally quoted return rate is 5% to 15% (and it varies by category). If Muse identified purchases are generating returns of 10% to 30%, then amazon is probably going to crack down on it. (Amazon Sellers pay from 30% to 50% for customer returns from their own pockets. https://medium.com/@myrealprofit/how-much-do-refunds-really-... )

llmslave 19 minutes ago | parent

Amazon cant hold this off for the long run, nobody wants to scroll through a list of two near identical packages of toilet paper. Agents will intermediate, and they will remove demand for low value items

gunalx 14 minutes ago | parent

Theh will probably consolidate more. Less choices even if they are worse is also easier for llms to do. Just look at how they pick technologies to use. Almost always the same for a given model.

meowtimemania 8 minutes ago | parent

I'm really enjoying muse+have used it to make some amazon orders. This is a big bummer to me, in future I'll find an alternative to amazon OR find an alternative to muse.

No matter what I'll continue to use some AI agent and some shopping provider.

causal 6 minutes ago | parent

Yeah I experimented with shopping and found it was really tedious jumping through the 2FA and other bot check hoops every third click. Agree it's a huge convenience if it would work though.

z0r 7 minutes ago | parent

I'm still annoyed that Amazon stopped sending order details in emails so that gmail can't show me when specific items are arriving. More recently 'getsauce.com' started to do the same thing, hiding order details behind an opaque link to their website. Every player in tech wants to force you to interact with their platforms purely on their own terms, to the detriment of the users.