digital_ins

Appalling: "Stealing is not corruption" Corruption involves kickbacks, bribes to allow for activities that would lead to an economic gain otherwise unachievable by those not participating in the corruption. Stealing involves achieving value by extraordinary means not condoned by law. Stealing may not be === corruption, but it definitely is == corruption. Anything else is a lie you tell yourself when on the receiving end of a kickback. And the author is very very naiive to imagine that corruption is chump change. Nigeria has vast reserves of oil but horrible infrastructure; it's pretty obvious that the corruption is ruinous out there.

I tried reading this article, but the horrible color scheme and irritating (parallax?) background just hurt too much :-(

This is reeeeaaaalllyyyy cool. Perhaps you should be targeting this at the college kids a whole lot more than the people on HN? (just a thought!)

Apart from this, could you elaborate a lil more on other use cases for a messenger which can hold conversations on a temporary basis?

I would really dig it if you could talk a little more about what this is. Without a description here or on the landing page, you're gonna slide off this list without making it to the front page of ShowHN

I'd be a wee bit careful: Nuance has a virtual assistant that they're calling Nuance. So you've gotta beware of trademark litigation (unless you're Nuance)

Yes! I was using it with encryption on. I was playing around with it a little bit.

The digital acoustic watermarking idea is awesome. Rather than have a weird buzzing noise, overlaying digital information over these codes is great.

If you've got enough traction, you should get a US corp built, because this is a trust-based product (right?). Having an OOO company is a little scary for US users.

Constructive criticism: A better soundtrack / voiceover is key. This one had too many things going on 2. You need to explain how the sound thing works. Is it frequency or amplitude based? (would help me understand what environment it would and would not work in) 3. I tried a demo and it sent the test code but showed me a different number on the screen.

Ah, when I asked about scalability for commercial usage, I meant - how do you plan to make enough money out of this to scale it. An automated bot that answers questions about me is a novelty, I'm curious about whether you would make enough money out of it to sustain it:: which would then make it worth my time to enter data about myself.

just tried to access the demo for the creator and it didn't work. Might help if you tested your links before uploading

great idea. How do you see this scaling for commercial usage?

it's quite obvious, really. Regular unauthorized access means: batten down the hatches, man battle stations, i.e. change password State-sponsored hacking means: delete social media profile, move residence, change phones, start praying, etc ;-)

Yeah, that's the stuff that happens when you make dodgy decisions based on how much of a maverick you think you are - and then look to invest in other 'seeming mavericks' like you.

Theranos is only the first, as more and more VC-backed companies come close to the border where private funding and public funding meet (commonly known as 'get ready for an IPO'), you will see a steep and shocking correction in valuations. Not a 'bubble bursting', but more of a 'okay, put the bong down, we gotta drive home now' sorta thing.

really disappointing that the app developer doesn't even condescend to replying to posts on his ShowHN posting

I like the idea, but I think that you'd really be able to convince a customer of the value proposition by including a nice demo of the stuff that's behind the 'sign up' page

I am appalled that the combination of Digital + Print costs as much or less than the digital subscription alone.

NYTimes, WSJ all seem to forget the basic principles of economics which dictate a certain amount of price elasticity is essential to bring in more customers. It seems almost punitive that digital customers pay at par with print customers when the incremental costs for the latter are relatively so minor.

If NYTimes is really serious about boosting it's pure digital subscriptions, scaling back the height of the paywall is key - not to a trivial non-zero sum, but to an amount where the pure digital customer doesn't feel ripped off.

This is a very cool swiss-army knife for the abundance of cab aggregators in India. Am gonna give it a whirl the next time I need transportation.

Because I've not used it yet, I'm curious as to whether you're using deep-linking to get to the other cabs? If the taxi companies offer APIs, I think another powerful use-case would be to link up to those APIs. Without them, this is a pretty good tool!

this really is something! One thing that I would love to see is font pairings that occur together frequently.

Being an NLP guy myself, I think this is a big BIG BIG deal and was really hoping that it would be Apple that would provide a way to hook into the voice recognition interface back when they were talking about their watch. :-(

Voice recognition has begun to come to our devices in the most unexpected way. The old way of thinking was that voice recognition would come via a powerful process running on a powerful PC, locally (remember Dragon Naturally Speaking?). In fact, if you look up many US patents re: voice recognition, you will see inventors who explicitly say that it's not possible to have voice recognition over a streaming connection.

Today, the easiest way to access free voice recognition is in your Chrome browser, with Google only very recently bringing Speech to Text to the Mac version (they've had Text To Speech for a while, I think). Firefox's bringing it out real soon and MS is... well... I dunno. They've got Cortana and Satya Nadella recently enjoyed using it so much that he probably doesn't want his users to have anything else :-) .

The unintended side-effect of this slow creeping of voice recognition tech into our devices is that it's gonna spawn a whole host of Medium articles on Design and I'm probably gonna have to subscribe to them to have some artist tell me how users would like to use my tech :-(

At the core of it - you're going to be leading engineers. A lot of how this leading happens depending on the organizational makeup of the firm you're going to. (This is super super important!!!).

So figure out the organizational culture before you join. Understanding this helps you understand the confines within which you will operate (for the first couple of months, atleast). If you're replacing someone, understand how they worked, what gaps were there in that person's performance and how you could cover those. If you're the first tech lead, then figure out what problems they're having that necessitated the creation of the tech lead position.

So I dunno if you've read Daniel Gilbert's (only) book - but he does kinda say (to paraphrase) that people get tougher when you bang them around.

The only danger is that some people don't. My school got the reputation of the suicide school of the north east because the nature of the school was so darned tough. Couple that with the unbelievably dull and cold winters and you had a situation where they had to put up fences over the bridges (making it even more depressing). Even the weak anti-suicide program they had was laughable.

This sweeping 'coddling' that schools are doing for their students can be put down to a lazy libertarian psychology or looked at for what it really is: that schools would rather have 16,000 wimps than have the stain of a kid take his own life

to what you're saying - I would think jogging would be a prime example of 'network' effect. If you head out to Beverly Hills, you don't see anyone jogging except the dog walkers. But in certain parts of say, Hyderabad in India, the super-rich areas have massive number of 'walkers' (even though the air is highly polluted and walking there is probably detrimental)

This sounds very cool. Any chance you could draw a flowchart diagram to help people (like me) who have no prior experience with AWS IoT? (like, which device is responsible for generating LAT-LONG, etc)

Actually, today's speech recognizers are really good and ones like Google's have accuracy >90%. So writing a speech recognizer ourselves with 50% accuracy and training it endlessly without ever reaching Google SR's accuracy sounded like a terrible 'reinvent-the-wheel' idea.

Instead, we wrote algorithms that further boosted the accuracy of the existing speech recognizer with machine-generated phonetic dictionaries. (You can think of it as standing on the shoulders of giants). The plan is to work with a couple of partners to train our machine learning system for their products to really bump up SR accuracy.

If you're building an app that's going to go live, send us a message, we'd be happy to work with you (for free!)

if you're curious to try this out - I'd recommend you sign up (free), download the IPA file (also for free) and try it out (also free!).

To help speed up your ios dev, we've also posted all the libraries you'd need on GitHub and on the first page of the Leova admin console, once you've signed in

So that's where Leova really shines. We saw this shortfall in accuracy when we were using iSpeech for testing on iOS, so we built a machine learning system that generated structured phonetic dictionaries.

This means that Leova provides a layer of error correction over and above specifically iSpeech's transcriptions. And with frequent usage (days, not weeks) you'll see Leova's interpretations of buggy iSpeech transcriptions really blow competing NLP parsers' resilience out of the water

Thanks! We've actually built our own probabilistic information extractor, so we don't use the POS tagging, so NLTK is sorta redundant for the approach we took (and also less accurate)

orrrrrr, if you watch the demo video really closely, you'll see how you can try the system out without actually having to sign up

Hi Everybody! I'm part of the team and we built Leova to make adding fully-functional natural language interfaces to your home-brewn or production-scale apps would be super easy. Integration is (supposed to be) super simple and deployment, even easier. We've got a couple of short videos on the website to show off how effortless it's supposed to be.

The NLP AI is conversational and as people use it, a small machine-learner learns to 'fix' errors in the recognition of whichever speech recognizer it has been paired with it. The entire system has been modeled around human conversation and works without keywords and without forcing users to speak to it in a rigid way, like an IVR system.

This first release is for LeovaTravel, which is this one :-)

The next 3 months will see us deploy the next two variants - a natural language AI for ordering food; and one for shopping / ecommerce, both wrapped in APIs.