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

Rabbit: LLM-First Mobile Phone

rabbit.tech

Post reply on HN
Print view View original

411–420 of 547 posts

Re: Rabbit: LLM-First Mobile Phone

#411

Post by danso »

dvt wrote: post #287

I feel like the R1 is getting a lot of weird hate here on HN. I get it, it's a bit pretentious and wants to be very Apple-like, yadda yadda. Probably not that much value generation, it's overpriced, and no one wants to carry another thing in their pocket. And yes, the company will likely go bankrupt (or pivot into something less exciting). It's also all in the cloud (no local inference) so it doesn't really even "do"…

> * For example, if I wanted to do a simple find-and-replace on my 2021 M1, I'd have to use something like grep (a tool originally written by Ken Thompson in 1973—a half century before my laptop was built) and look up flags and syntax I inevitably constantly forget. We should have been past this at least 15 years ago.*

What does the oldness (and long usefulness) of grep have to do with the problem you described, which is the very hard problem of translating natural human speech into machine commands?

Re: Rabbit: LLM-First Mobile Phone

#412

Post by JKCalhoun »

ergocoder wrote: post #410

I'll call it. This is vaporware. This is similar to Fast and Bolt. 2 reasons: - It will be better as a phone. People will not use it and the initial hype will die quickly. - It is seemingly useful but useless in practice because many things are impossible to do. The demo is already so fake. With the two above reasons, the momentum and funding will dry up quickly. Check back in a year.

At the same time, I don't see why the Apple Watch, just as an example, won't go this direction.

Come too think of it ... maybe that's the plan: to be bought up by a big tech company.

Re: Rabbit: LLM-First Mobile Phone

#413

Post by abrichr »

Darthy wrote: post #298

In the linked keynote, Jesse Lyu mentions that LLM won't help us actually do tasks - there are currently no so-called "agents" that do something simple like book a flight - the best way to do it is still to click the buttons yourself. Rabbit means to solve that by creating a "LAM", a "Large Action model", which is a service by Rabbit that will click interfaces for you. I'm not sure this is the right approach - if it…

> Rabbit means to solve that by creating a "LAM", a "Large Action model", which is a service by Rabbit that will click interfaces for you.

https://openadapt.ai is an open source app that runs on your local machine that clicks interfaces for you -— but only for repetitive tasks that you show it how to do.

Re: Rabbit: LLM-First Mobile Phone

#414

Post by abrichr »

ryanSrich wrote: post #366
jonplackett wrote: post #307

Earlier quoted context omitted.

The idea that anyone is going to book a vacation this way is ridiculous. Vacations take extensive thought. There’s no way a voice interface would work for this even if it was the Star Trek computer.

No one ever shows what these should and could actually be used for - incredibly boring and repetitive tasks. "Hey rabbit, respond to the support ticket you just told me about with XYZ". "Hey rabbit, find all of the sales meetings I had last week, cross reference the emails of the attendees with existing emails threads, and then generate follow ups for each. Keep them short, to the point, and read them back to me befo…

> No one ever shows what these should and could actually be used for - incredibly boring and repetitive tasks.

Agreed! That’s precisely why we created https://openadapt.ai.

Re: Rabbit: LLM-First Mobile Phone

#415

Post by joezydeco »

underyx wrote: post #183

Two things in the keynote that seem faked: 1. at 13:13 there's a demo ordering a ride 'to home', then the user requests a car change to fit six people, and what's shown as a seamless switch from UberX to UberXL also updates the destination from the home address to LAX airport. 2. at 14:05 the device confirms and recites a pizza order, but the screen displays "chesse" with a typo while the voice reads out "cheese", so…

Is there any text-to-speech engine that would pronounce Cillian Murphy's name correctly on the first try?

Re: Rabbit: LLM-First Mobile Phone

#416

Post by shermantanktop »

joezydeco wrote: post #415
underyx wrote: post #183

Two things in the keynote that seem faked: 1. at 13:13 there's a demo ordering a ride 'to home', then the user requests a car change to fit six people, and what's shown as a seamless switch from UberX to UberXL also updates the destination from the home address to LAX airport. 2. at 14:05 the device confirms and recites a pizza order, but the screen displays "chesse" with a typo while the voice reads out "cheese", so…

Is there any text-to-speech engine that would pronounce Cillian Murphy's name correctly on the first try?

It’d be doing better than most people if it could.

Re: Rabbit: LLM-First Mobile Phone

#420

Post by anentropic »

yreg wrote: post #398
Calamityjanitor wrote: post #238

This is exactly what Siri wanted to be. Compare with the original Siri keynote https://vimeo.com/5424527 I can't find the exact ~2010 article from before being bought by Apple. I remember in an interview they were talking about making a web agent that could operate and perform tasks on any website, to avoid being locked out by APIs.

I'm very interested in what Apple does with LLMs on iDevices. They have the right hardware for it and they have all the motivation, with their focus on on-device processing. OTOH they also have a pretty bad history with their AI assistant.

They should have been building this instead of Vision Pro white elephant
Post reply on HN