Maybe I'm not the exact target market, but I always carefully pick my search parameters and research my decision making. For showing up, this is impressive to send flowers, but there's always some minor details that break the experience, like the shop is closed, I can't pay via X service. The Hotels.com booking is great, but not everybody is that rich to book a villa :)<p>As I see it, Viv aims to be the Google of service providers which is a pretty neat goal, nobody did that successfully, so I'm rooting for them. The platform approach is also a good marketing strategy that they want to allow it to embedable. For example letting people say ask basic things about a monument in a remote national park. The applications are really limitless if it fits on a Raspberry PI like card.<p>As someone who like to break things, please Viv process this for me:<p>"VIV, please find me a cheap flight to Vienna, check Ryanair and Wizzair on the 21st of may for two adults"
"order the results by price"
"show details for the second one"
"show the previous one"
"are luggage included in the price?"
"book it, but skip all marketing offers from the company"
"find AirBnB apartments for that day that has wireless internet"
"order results by price"
"ok, then search in the range of 10 to 50€ per night"
"select the first place"
"message host: looking forward to meet you"
"book it."
"Viv, bring me a beer"
Impressive demo, but yet i can't help thinking that the "big" question still remain largely unsolved for those services.<p>Two points :<p>1/ viv keeps mentioning a "breakthrough" in computer science because they managed to create a program that will scan a network of services and fill the parameters from the query. Now, those guys aren't jokers, so i presume there is definitely something fantastic behind, yet i can't help thinking that once you've "understood" the intent, it sounds quite close to a sql / graphql query planner.<p>2/ Which leads to the big question : guessing the intent. Every time i hear people mention "intelligence" or "understanding", i show them this : "hey siri, please do NOT set a timer for 4 minutes, i beg you". The fact is, it's just a trick. It recognize words, but it doesn't understand anything, it doesn't have concepts, knowledge or any experience. It's a dumb program that never had any life or sense to understand anything.<p>If i say "find me a good ticket for Chicago tonight". How will it know that i'm talking about the rock band that's playing tonight in Paris, France, since there's absolutely not a chance that a human being asks for a plane ticket for such a big trip just a few hours in advance ?<p>This is, to me, <i>the</i> big and interesting question that online assistants makers needs to solve. Now of course, viv is aiming for a product release this year, so they're building an intermediate solution, where developers will manually associate keywords to services, in an "easy" way. And it will sort of work.<p>Yet i'm still waiting for the real "big" advance.
That's pretty impressive!<p>Maybe this isn't as big a concern as I think, but the ease of these commercial demos relies on a crucial assumption: that I am nearly completely price insensitive. I'm not a cheapskate, but I don't usually purchase the first product a vendor puts up because there's usually something better and slightly less expensive if you put in an additional minute or two into the search. But with the demos given (ordering flowers, booking hotels, etc.) it's usually a sample of 2-3 different options, and from what I've seen, usually higher end items.<p>Will such assistants still maintain their usefulness when I don't have the monetary pleasure and freedom of just saying "Yes" to the first option offered?
EDIT: added intent in the conversation<p>I think one thing missing from these interactions is the idea of conversation. Let me elaborate. Distilling complex tasks into a single query yields complex queries. Drill down with simple initial query requires manual input from provided results. If there was some short term memory as part of the conversation we could potentially arrive at a much clearer intent.<p>e.g. "Viv I want to go to San Diego this weekend. Find me some cheap flights." (Travel, San Diego, Saturday, air fare)<p>"No problem. How long are you planning to stay?" (return trip)<p>"I want to come back on Sunday night" (return trip Sunday night)<p>"Will you be requiring accommodations?" (air fare, hotel combo lookup)<p>"Yes I want to stay near the waterfront" (area for hotel, duration 1 night based on previous reply)<p>"Sure no problem. Will you require a car rental?" (air fare, hotel, car combo lookup)<p>"No that's fine" (air fare, hotel lookup only)<p>"Ok here are some good deals for flights and hotels on the waterfront for Saturday"<p>"What are my options for a morning flight?" (Morning flight on Saturday)<p>"These flights are in the morning" (Subset of the flights found earlier)<p>"How about late Friday flights?" (Friday evening flights to San Diego)<p>"Here are your options for a late Friday departure with the same hotel options" (new departure schedule but same hotel options, new duration of 2 nights of hotel stay)<p>"Show me just the flights for Friday"<p>"Here are just the flight options for Friday"<p>"Nah let's see the Saturday options again" (go back to previous criteria)<p>"Here are the Saturday flight and hotel options"
"This is software writing itself". Welcome to the 70's.<p>The speaker comes across as poor in my opinion.<p>He's doing marketing, not informing the audience, and he's not even that interesting.<p>He's saying a ton of stuff which is completely unsupported by reality.<p>Assistants are the next paradigm shift? Have you tried to use Siri in a car? Or at all? It's a joke.<p>Weather examples? Show me something I use more than once a week.<p>Show me something actually <i>hard</i> that requires intelligence. Show me a recommendation service which does not suck. Show me an assistant that figures out something nice to do this weekend, based on location inferred by my plane tickets, weather, my taste, and what's available.<p>Show me an assistant that <i>shops for the cheapest flowers</i>, not force me to use the service of the app maker's choice.<p>Please, let's stop the "wow I can speak to my telephone, and it tells me the weather" bore. Please.
So, is this an island parser on top of a Bayes net?<p>Haven't read the papers for a few decades, but I wonder how the auto generated code differs from General Problem Solver: <a href="https://en.wikipedia.org/wiki/General_Problem_Solver" rel="nofollow">https://en.wikipedia.org/wiki/General_Problem_Solver</a><p>Hope the above doesn't sound dismissive. Was working at company incubated at SRI. We were hand coding similar workflows. My guess is that VIV is a 10x improvement in workflow.<p>Be that as it may, it still seems to be constrained by Humans creating ontologies. So, may suffer scaling issues. Somewhat akin to Yahoo's Human curated websites vs Google's auto generated backrub.
Sort of concerned about this:<p>Who decides which service to use for what thing? What if I don't want to use uber for my cab but a new uber like service that my friend just started? If my friend adds an experience to viv for his service, how does viv decide which one to use? what about neutrality etc.?
You can make your own "PowerShiri" in Powershell, by tapping into the .NET Speech Recognition library [0].<p>Having built out a novel bot this weekend [1], this became an obvious next step. The PowerShiri demo linked lets you set up a key/value table, and when Speech Recognition matches a key, it executes the Powershell command held in the value. In this way, you can issue the same command, e.g. "What is the time?" - and receive a dynamic response. The voice output is generated by the ubiquitous Out-Speech function.<p>Dynamic program generation, more like dynamic query generation, is the obvious next step. Instead of generating a massive key/value table for every possible combination of commands, you treat each word as a key, then have each value generate the next set of keys.<p>It's both annoying and gratifying to learn a weekend project is facing the same technical challenges as a well-funded corporation.<p>[0] <a href="http://stackoverflow.com/questions/9361594/powershell-can-speak-but-can-it-write-if-i-speak" rel="nofollow">http://stackoverflow.com/questions/9361594/powershell-can-sp...</a><p>[1] <a href="https://github.com/Gilgamech/PowerShiri/" rel="nofollow">https://github.com/Gilgamech/PowerShiri/</a>
>The live demo went off without any major glitches<p>At 3:54 - "The other thing I want to note is that Siri... err ummm Viv's audible voice is something we're still working on."
Who decides what "app" answers when I want to get flowers? I bet there are tens of apps doing it currently, and users have their own preferences. Same for anything weather-related. What would be my chances as a developer to compete with the more mainstream providers? While it isn't huge now, it seems impossible with this future, where the user practically doesn't even chooses the tools she wants to use.<p>(sorry if the question was answered in the video and I missed it)
Sounds similar to IBM Watson that generates Prolog code on-the-fly. I wonder if Viv is using Prolog behind the scenes too (behind the visual AST representation).
Well, booking hotels <i>their</i> way has always been easy. Just open the website, see the first three search results and book one which 'looks' the best in photo (assuming you have all the money of the world). In reality, it's not that easy, there are so many parameters to consider which they haven't. Price, smoking preference, number and size of beds, cleanliness, wi-fi and its availability and speed, other amenities, parking and so much more you have to consider before deciding to get into a hotel. This personal assistant doesn't seem to do any of that. The problem is, I don't even know how can we automate or predict that because these are very personal preferences and change with each trip.
Great to see they finally unveiled Viv. They've been working in stealth for a few years. Six months ago, I met the team for interview & offered a position as Mobile Architect. I declined the job for reasons unrelated to Viv. I met the technical cofounders, Adam Cheyer & Chris Brigham. It was just awesome to talking bunch of guys who made original Siri, and personally worked with Steve Jobs. You can read up on Adam, not only he's an AI God, but he's a super nice guy. If I were to bet on one AI company, they would be it. Like Kittaus eluded to, it's going to take some time to see what use cases ordinary people use for but it will be exciting. Best of luck to them.
I can't help but feel bearish on any company like this that tries to exist outside of either the android or iOS integration. It's a desert without OK Google/Siri
This is interesting to me because it really pulls together a number of disparate technologies to get towards the goal of being "the AI". There's nothing he talks about that's overly complicated on it's own, but integrating them all together is the key.<p>Once you have the voice to text (ASR) from Nuance (or any of the other companies in this space), then it's a question of properly recognizing the contextual intent. Not a trivial task, but certainly getting easier with technology available. I think the visual models displayed in the demo are fairly telling as to how they handle this. Keywords for a domain (e.g. weather) become useful to determining the intent of the speech, as well as variables used in queries, or the data returned for display to the user.<p>For example, if I ask a question that mentions temperature, and a time, but not a length of time, it's fairly obvious I'm asking for something in a weather domain rather than a recipe domain (and vice versa).<p>I'd guess the developer integrations he mentions then become a matter of defining those data points/variables for the different data source so the "AI" can build the application to execute.<p>It's like an enhanced API integration model. You need to know all of the input/output parameters to integrate with an API. The intent/contextual piece also uses the individual data points for the contextual intent recognition in the voice-to-text area.<p>The other interesting aspect is actually storing the preferences for the commerce side of things. Airlines do this already for saving preference of aisle vs window seats. They're taking things a step further to remember those types of "qualifying data" for interactions you have so they can be saved across areas (read: API calls).<p>I suspect there will be a ton of work on behalf of the APIs too handle data in this way too, that's what he says when it will take some time to see the direction things go in. If I opened up my existing airline travel APIs to this today, it's unlikely anyone could correctly interact in a way that would provide all of the information needed to actually book a ticket. So, there will need to be some back-and-forth communication of those missing items. If someone finds a ticket they want to book and says "order it", then viv will need a way for the API to communicate "That's great, but you also need to provide your TSA known traveler ID number". Then, because it's something an API has asked for, Viv will know it's a data point it should save for later.<p>Let's hope Viv has un-breakable encryption and security with all of those "personal preferences"...
I pitched a very early similar idea of Playa (<a href="http://getplaya.com/" rel="nofollow">http://getplaya.com/</a>) at Launch IoT Conference in front of the audience and was literally met with Jason and the panel of judges who had no idea what I was talking about.<p>After the pitch, I had representatives from NTT Docomo, Cisco, GE, and a few others come up to me and were interested to chat about ambient intelligence and stay in touch.<p>This is telling of how things are now changing and moving towards the Ambient Intelligent future.<p>We're still building towards this future, and would like to meet up with anyone in SF that wants to help out.<p>We want to put souls inside your devices. :)<p>[EDIT] Luckily, Robert Scoble was kind enough to give us some outstanding attention on his feed, and also hook us up with RackSpace's startup partnership program. Huge thank you to him for that.<p>[EDIT] We just set-up a baqqer account to be as transparent as possible as we build out the product. <a href="https://baqqer.com/projects/playa" rel="nofollow">https://baqqer.com/projects/playa</a>
Am I missing something, or was the demo on the Mac just a pre-recorded QuickTime video? The menu title bar said "QuickTime Player" and the title of the window was "Movie Recording".<p>I don't doubt it works for real, but isn't it a bit weird he didn't acknowledge the "live demo" was actually pre-recorded?
The developer experience looks interesting, looks like some kind of flow based, quite restricted coding environment? Getting that ecosystem and incentives for devs right will be the key to their success.<p>I've been using Amazon Alexa for a month and really struggling to find 3rd party apps or "skills" for it that I would want to use. Also, namespace pollution quickly becomes an issue when you start having thousands of apps that you didn't install but are there to use.<p>I think Google will dominate the voice assistant space soon as they have so many crucial services already there.<p>Good to have competition though!
Any ideas on what is used to power the graph demo used about 9 minutes in (<a href="https://youtu.be/MI07aeZqeco?t=539" rel="nofollow">https://youtu.be/MI07aeZqeco?t=539</a>)? I haven't seen something so fast and seamless in a browser.
Live demo starts at 9:00. I am very very impressed.<p>One of my top complaints with Siri is inability to chain commands. If you want to modify a command you have to start over from scratch. Which often results in it misunderstanding another word. It's very, very frustrating.
I always cringe when someone says 'software writing itself'.
Looks to me more like software constructing a query, much like an SQL query. Not sure that is a 'computer science breakthrough'
Still, the whole thing is pretty neat.
I live by myself so the idea of talking when another human is not around or hearing my voice is sort of offputting. So cool to see this technology evolve, but I hardly can put Siri (an amazing technology) to good use!