The shopping assistant plugs into browsers like Chrome and Safari to compare prices and surface deals across tens of thousands of retail and resale sites in real time. It essentially serves as your own personal deal finder: Say you’re looking at a $200 dress from Anthropologie, Phia can find and compare prices at secondhand sellers to help customers find a better price.
Gates and Kianni first brainstormed startup ideas in their Stanford dorm room, cycling through concepts before landing on a consumer tool that included Gates’ interest in women’s empowerment (likely modeled after her own mother) and Kianni’s sustainability focus.
I don’t think a coupon tool that wastes excessive resources is either empowering or sustainable.
I’d be surprised if a human were behind it. This is exactly the kind of thing that can be vibe coded pretty fast and is mostly just reselling fancy Google searches through an LLM. I did a quick skim of the website and it’s just a bunch of items scraped from big brands with lots of similar looking images of other products. There’s too many sites for me to really believe they’ve made integrations with all of them.
The insane valuation is because of her name not because the tech is good. The only way to make money on this is the customer data. The margin on that is going to be fucking minuscule especially once LLM costs start going up so they can make money. This adds nothing of value on top so it will go away almost immediately.
Ohhhhhh yeah this is just 100% LLM garbage. No machine learning. Forgive me; I don’t often talk to people that understand the difference so my default is human vs LLM not LLM vs <insert something AI that isn’t LLM>. Trying to explain machine learning vs LLM summaries to data company business executives wears you down.
Edit: also my frame of mind was on the more recent Waymo “our AI just goes to a human when it’s hard” news that followed the Amazon “it was always a Mechanical Turk” news which is why I jumped to human.
I’d be concerned about monied motives driving suggestions. Even if it starts out neutral, how can we as consumers be sure it won’t become corrupted? Enshittification is par for the course these days, I’d be extremely wary about relying on an app to tell me real, unbiased price info unless its mechanisms and sources are (and remain) completely transparent.
I don’t think a coupon tool that wastes excessive resources is either empowering or sustainable.
I’d be surprised if there’s actual AI behind it. That functionality already exists and works just fine.
it’s a small indian man named Al.
I’d be surprised if a human were behind it. This is exactly the kind of thing that can be vibe coded pretty fast and is mostly just reselling fancy Google searches through an LLM. I did a quick skim of the website and it’s just a bunch of items scraped from big brands with lots of similar looking images of other products. There’s too many sites for me to really believe they’ve made integrations with all of them.
The insane valuation is because of her name not because the tech is good. The only way to make money on this is the customer data. The margin on that is going to be fucking minuscule especially once LLM costs start going up so they can make money. This adds nothing of value on top so it will go away almost immediately.
Oh, I didn’t say a human, I just meant a standard non-AI algorithm
Ohhhhhh yeah this is just 100% LLM garbage. No machine learning. Forgive me; I don’t often talk to people that understand the difference so my default is human vs LLM not LLM vs <insert something AI that isn’t LLM>. Trying to explain machine learning vs LLM summaries to data company business executives wears you down.
Edit: also my frame of mind was on the more recent Waymo “our AI just goes to a human when it’s hard” news that followed the Amazon “it was always a Mechanical Turk” news which is why I jumped to human.
I’d be concerned about monied motives driving suggestions. Even if it starts out neutral, how can we as consumers be sure it won’t become corrupted? Enshittification is par for the course these days, I’d be extremely wary about relying on an app to tell me real, unbiased price info unless its mechanisms and sources are (and remain) completely transparent.