Wildfire was doing much of this twenty years ago.[1][2] Listen to their demo. That system was way ahead of its time. It was quite popular, but compute intensive enough that the service was expensive. At its peak, Wildfire was a standard offering on Orange cellular service in the UK, but they dropped it in 2005. Somewhere in the 2000s, Microsoft bought the technology and trashed the product, although Cortana is to some extent a successor. Now a small startup has the original technology, and offers it, at an excessive price.
One key to doing this is that the system must consistently respond to the user in about 200-300ms. Long delays, as with Google voice dialing, are unacceptable. That was why it was so expensive in the 1990s; the company had to provision a lot of server capacity per user. Today, it ought to be much cheaper.
$49 a month reflects what Wildfire cost in the 1990s. Today, it should cost about $2/month.
This looks awesome. I use reminders in Inbox with reckless abandon, love them. Please tell me there is an open source version of this or that I am missing something in gmail to do this. Pretty please.
That said: There is no way I am spending $49 a month on this. There is a point where you have to analyze how many $5/$10/$15 dollar a month subscriptions you have. Pretty soon it can escalate to the cost of a monthly vehicle payment or the hiring of a part-time employee. Furthermore, can get a human to do this for $10 an hour or so using a virtual assistant?
If I was super-crazy-busy and making ridiculous $$$, sure. But as a cash flow restricted mortal with a mortgage a full time job and a side project... $49 in addition to the myriad of services/equipment I require (accounting, marketing, materials, equipment maintenance, etc) is not going to happen.
An intelligent assistant is fantastically useful. I sometimes pay $50/month (depending on how much work she does) and she is worth every penny.
E.g., "Deepika, call up brokers and schedule me apartment viewings in $X location for $Y or less, schedule these appointments on monday evening back to back. Do the same for $Z location on tuesday evening."
Of course, Deepika (name changed a little) is actually a real girl; I can only afford her because her choice of birth country allows me to pay less than US minimum wage.
In my experience these robot assistance are not as smart as Deepika. Scheduling in particular is a use case designed to make the user feel important rather than actually save people time. Doodle ( http://doodle.com/ ) substitutes UI for AI, but it works brilliantly. Every email bot I've been the victim of just fobs work off from the user to their counterparty.
Exactly what I was thinking. If you read the end of TFA you'll see that the main benefit here is that they provide a human force to curate it, etc. That means that even if they open sourced it today it won't do you and me any good without the human element. Then again, if there is an OSS version of this that someone knows of, please do point it out.
Having said that the price is too high for me personally. I'd consider $5/month. Also since I schedule way fewer meetings than the author of TFA, I would need longer than a month to try it out.
> There is no way I am spending $49 a month on this.
You're not the target market then. For that matter, neither am I. The target market is people who need an admin assistant, but can't pay the five figures per year yet.
My first thought was "ow", but thinking about it more, that's not a bad price point to filter out people who think "oh, that's neat" rather than "oh, thank God, I really need something like that".
Small nitpick, but i would lose the first GIF which has about 6 emails scrolling by in quick succession. I'm not a speed-reader, so even managing to make out the grey summaries at the top took me like 6+ iterations. That's pretty frustrating, and i only stuck at it because i'm procrastinating and should actually be doing the dishes, because i should actually be changing a light bulb, because i should actually be rinsing my yak, because i should actually be writing my thesis manuscript. In that order.
EDIT: or at least slow it down, because of course the point it's trying to make is cool and/or worth it and/or noble.
Some nice PR for Clara, congrats to the Clara team!
I observed this space for some time and came to following conclusion (from an investment perspective):
- There are tons of competitors, some of them very well funded such as x.ai
- Assuming that at some point, let's say 18-24 months the AI of all providers is near perfection, user should be able to just switch from one provider to the next one => where is the lock in? Maybe it's learning the user's preference and behavior?
- At the moment the AI is damn impressive but at the same time it's easy to confuse the tech; just bounce some emails back and fourth with "Hey let's meet today", "Where?", "At the Ritz, 3pm", "Ok", "hey I have to postpone, will come back to you with a new time", "no problem", "tomorrow is better, same location, same time, but let's meet directly at the restaurant"; semi-automatic systems (i think Clara is) which use human operators to do final checks might deal with such messy conversations but then it's the question about scaling again
- You can't include the bots in messenger conversations (minor)
- Having a real PA is the key to success; the better she knows you the more she will take of your shoulders. The PA very often acts as a proxy for you and sometimes it's wise to use this proxy and sometimes not—depending on the other party and their status. It's perceived as a status symbol and it sends the signal 'this person is very busy', this is in particular true in a hierarchical context like in a corporation where only higher levels have dedicated PAs. It's also common not to approach higher status people (compared to your own status) with your PA. However, having an AI PA feels just cheap, like "Hey, I am busy but I can't afford a real PA", this is the main problem, if you are really important und if you really run a successful business you NEED to have a ded PA, no AI PA will be able to do the job of a real PA; important people have two, three dozen meetings a week and just for scheduling and rescheduling you need a ded PA, less important people have so few meetings a week that they can schedule them themselves; maybe this perception will change in 10 years where AI PAs get better and better and everybody has one but the status issue still stays
Would love to know your opinion? Is this a field to invest?
I used that! The best thing was that it was free IIRC. Also, does anyone else find it weird that all these personal assistants are created with a female persona? Just wondering if there is a broader thing here.
[+] [-] Animats|10 years ago|reply
One key to doing this is that the system must consistently respond to the user in about 200-300ms. Long delays, as with Google voice dialing, are unacceptable. That was why it was so expensive in the 1990s; the company had to provision a lot of server capacity per user. Today, it ought to be much cheaper.
$49 a month reflects what Wildfire cost in the 1990s. Today, it should cost about $2/month.
[1] https://www.wildfirevirtualassistant.com/ [2] https://en.wikipedia.org/wiki/Wildfire_Communications
[+] [-] mmatty|10 years ago|reply
[+] [-] witty_username|10 years ago|reply
[+] [-] monkmartinez|10 years ago|reply
That said: There is no way I am spending $49 a month on this. There is a point where you have to analyze how many $5/$10/$15 dollar a month subscriptions you have. Pretty soon it can escalate to the cost of a monthly vehicle payment or the hiring of a part-time employee. Furthermore, can get a human to do this for $10 an hour or so using a virtual assistant?
If I was super-crazy-busy and making ridiculous $$$, sure. But as a cash flow restricted mortal with a mortgage a full time job and a side project... $49 in addition to the myriad of services/equipment I require (accounting, marketing, materials, equipment maintenance, etc) is not going to happen.
[+] [-] yummyfajitas|10 years ago|reply
E.g., "Deepika, call up brokers and schedule me apartment viewings in $X location for $Y or less, schedule these appointments on monday evening back to back. Do the same for $Z location on tuesday evening."
Of course, Deepika (name changed a little) is actually a real girl; I can only afford her because her choice of birth country allows me to pay less than US minimum wage.
In my experience these robot assistance are not as smart as Deepika. Scheduling in particular is a use case designed to make the user feel important rather than actually save people time. Doodle ( http://doodle.com/ ) substitutes UI for AI, but it works brilliantly. Every email bot I've been the victim of just fobs work off from the user to their counterparty.
[+] [-] IgorPartola|10 years ago|reply
Having said that the price is too high for me personally. I'd consider $5/month. Also since I schedule way fewer meetings than the author of TFA, I would need longer than a month to try it out.
[+] [-] pnathan|10 years ago|reply
You're not the target market then. For that matter, neither am I. The target market is people who need an admin assistant, but can't pay the five figures per year yet.
[+] [-] mcdoug|10 years ago|reply
There's also https://www.indiegogo.com/projects/cubic-your-personal-ai-wi....
And finally there's Facebook M.
[+] [-] zyxley|10 years ago|reply
My first thought was "ow", but thinking about it more, that's not a bad price point to filter out people who think "oh, that's neat" rather than "oh, thank God, I really need something like that".
[+] [-] toothbrush|10 years ago|reply
EDIT: or at least slow it down, because of course the point it's trying to make is cool and/or worth it and/or noble.
[+] [-] glossyscr|10 years ago|reply
I observed this space for some time and came to following conclusion (from an investment perspective):
- There are tons of competitors, some of them very well funded such as x.ai
- Assuming that at some point, let's say 18-24 months the AI of all providers is near perfection, user should be able to just switch from one provider to the next one => where is the lock in? Maybe it's learning the user's preference and behavior?
- At the moment the AI is damn impressive but at the same time it's easy to confuse the tech; just bounce some emails back and fourth with "Hey let's meet today", "Where?", "At the Ritz, 3pm", "Ok", "hey I have to postpone, will come back to you with a new time", "no problem", "tomorrow is better, same location, same time, but let's meet directly at the restaurant"; semi-automatic systems (i think Clara is) which use human operators to do final checks might deal with such messy conversations but then it's the question about scaling again
- You can't include the bots in messenger conversations (minor)
- Having a real PA is the key to success; the better she knows you the more she will take of your shoulders. The PA very often acts as a proxy for you and sometimes it's wise to use this proxy and sometimes not—depending on the other party and their status. It's perceived as a status symbol and it sends the signal 'this person is very busy', this is in particular true in a hierarchical context like in a corporation where only higher levels have dedicated PAs. It's also common not to approach higher status people (compared to your own status) with your PA. However, having an AI PA feels just cheap, like "Hey, I am busy but I can't afford a real PA", this is the main problem, if you are really important und if you really run a successful business you NEED to have a ded PA, no AI PA will be able to do the job of a real PA; important people have two, three dozen meetings a week and just for scheduling and rescheduling you need a ded PA, less important people have so few meetings a week that they can schedule them themselves; maybe this perception will change in 10 years where AI PAs get better and better and everybody has one but the status issue still stays
Would love to know your opinion? Is this a field to invest?
[+] [-] HarrietJones|10 years ago|reply
http://lifehacker.com/321644/sandys-your-personal-assistant-...
[+] [-] IgorPartola|10 years ago|reply
[+] [-] cfontes|10 years ago|reply
"Clara, find me a time to do xxx next week"
That integrated with Google inbox (the remind me later feature) would be great!
P.S: I found it kind of expensive, isn't it?
[+] [-] cryoshon|10 years ago|reply