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

Airbnb CEO says it's still too early for AI trip planning

techcrunch.com

Post reply on HN
Print view View original

11–20 of 60 posts

Re: Airbnb CEO says it's still too early for AI trip planning

#11

Post by throw098320923 »

suddenlybananas wrote: post #9
throw098320923 wrote: post #7

AI works just fine for trip planning. I use it to scrape Airbnb apartments descriptions and reviews. Common problems such cockroaches or slow internet, are not usually mentioned directly, but with indirect special wording, and LLM can find those very well. Also scraping newest social posts and Tiktok videos is great way to find cool stuff.

I find it hard to see the advantage of scraping reviews versus just looking at the actual score.

Score is rigged, description does not match reality, negative reviews are deleted.

AI is perfect to answer questions like:

- is there a big table on pictures (proper workspace)

- is there a direct sea view

- any mentions of drain issues...

Re: Airbnb CEO says it's still too early for AI trip planning

#12

Post by soulofmischief »

ElectRabbit wrote: post #2

LLMs are stellar for trip planning. They can spit out a lot of potential options for a country or region. I just add all the suggestions to a map and then plan way around them. Also doing some manual research like opening times. Why I love this? I really don't want to research too deep into my destinations as this kills a lot of the magic seeing seeing them in person.

Plenty of positives, but consider the negatives with current-generation language models. It takes one hallucination to put your life in danger.

Re: Airbnb CEO says it's still too early for AI trip planning

#13

Post by nycdatasci »

I agree with this take. I tried to use OpenAI Operator to find a hotel a few times recently and every time it has made enormous errors like booking for a month and 3 days instead of 3 days because the end date was selected from the next month instead of the current one in the date selector. Completely unusable.

Re: Airbnb CEO says it's still too early for AI trip planning

#14

Post by nycdatasci »

throw098320923 wrote: post #10
rainonmoon wrote: post #8

Earlier quoted context omitted.

This is very interesting, can you share a bit more about your methodology/setup?

Android app, USB dongle with HDMI capture card. Screenshot goes to OCR and LLM. There is a virtual USB HID mouse build on Arduino that clicks on links.

For more on this setup, see “Web Scraping Fundamentals” by Rube Goldberg.

Re: Airbnb CEO says it's still too early for AI trip planning

#15

Post by foldr »

suddenlybananas wrote: post #9
throw098320923 wrote: post #7

AI works just fine for trip planning. I use it to scrape Airbnb apartments descriptions and reviews. Common problems such cockroaches or slow internet, are not usually mentioned directly, but with indirect special wording, and LLM can find those very well. Also scraping newest social posts and Tiktok videos is great way to find cool stuff.

I find it hard to see the advantage of scraping reviews versus just looking at the actual score.

I find that the only thing that works with Airnbnb is just to read all the reviews after excluding anything that isn't close to a 5.0 rating. Some kinds of negative comments in reviews can safely be ignored (e.g. the people who rented an apartment in the middle of downtown and were surprised that it's noisy), but almost any substantial negative comment in any review should be a red flag, no matter how many positive reviews there are.

Re: Airbnb CEO says it's still too early for AI trip planning

#16

Post by CharlieDigital »

throw098320923 wrote: post #11
suddenlybananas wrote: post #9

Earlier quoted context omitted.

I find it hard to see the advantage of scraping reviews versus just looking at the actual score.

Score is rigged, description does not match reality, negative reviews are deleted. AI is perfect to answer questions like: - is there a big table on pictures (proper workspace) - is there a direct sea view - any mentions of drain issues...

Only if you know what to ask.

You'll miss all the things you didn't ask about -- good and bad.

Re: Airbnb CEO says it's still too early for AI trip planning

#17

Post by eudhxhdhsb32 »

throw098320923 wrote: post #10
rainonmoon wrote: post #8

Earlier quoted context omitted.

This is very interesting, can you share a bit more about your methodology/setup?

Android app, USB dongle with HDMI capture card. Screenshot goes to OCR and LLM. There is a virtual USB HID mouse build on Arduino that clicks on links.

Wouldn't it be easier to just interact directly with their web interface through an automated browser?

Re: Airbnb CEO says it's still too early for AI trip planning

#18

Post by troupo »

foldr wrote: post #15
suddenlybananas wrote: post #9

Earlier quoted context omitted.

I find it hard to see the advantage of scraping reviews versus just looking at the actual score.

I find that the only thing that works with Airnbnb is just to read all the reviews after excluding anything that isn't close to a 5.0 rating. Some kinds of negative comments in reviews can safely be ignored (e.g. the people who rented an apartment in the middle of downtown and were surprised that it's noisy), but almost any substantial negative comment in any review should be a red flag, no matter how many positive r…

for both Booking and AirBnB make sure to read recent reviews. Booking especially likes to default to "relevant" reviews (which are always glowing reviews from five years ago)

Re: Airbnb CEO says it's still too early for AI trip planning

#19

Post by troupo »

ElectRabbit wrote: post #2

LLMs are stellar for trip planning. They can spit out a lot of potential options for a country or region. I just add all the suggestions to a map and then plan way around them. Also doing some manual research like opening times. Why I love this? I really don't want to research too deep into my destinations as this kills a lot of the magic seeing seeing them in person.

> They can spit out a lot of potential options for a country or region.

If you're going to all the crowded tourist-y places, sure. It's a struggle to make them suggest anything off the beaten path for obvious reasons

Re: Airbnb CEO says it's still too early for AI trip planning

#20

Post by CharlieDigital »

ElectRabbit wrote: post #2

LLMs are stellar for trip planning. They can spit out a lot of potential options for a country or region. I just add all the suggestions to a map and then plan way around them. Also doing some manual research like opening times. Why I love this? I really don't want to research too deep into my destinations as this kills a lot of the magic seeing seeing them in person.

The founder of Viator had an interesting take on this in an interview (How I Built This [0]).

The gist is that the people who deeply research and plan, they create this perfect vision of the trip in their head. When you actually go on the trip, things can go wrong; weather is bad, destination was crowded, you got sick, etc. But years later, the perfect version of the trip and the actual version of the trip get mixed up in your memories. People who didn't plan, didn't have this "idealized" version imprinted.

I find this to be true so my spouse and I plan our trips meticulously and research everything.

We also capture a detailed log of our trips. And I think this combination really helps to highlight and reinforce the good memories while letting the not so good ones fade.

[0] https://wondery.com/shows/how-i-built-this/episode/10386-via...

Post reply on HN