Hacker Newsnew | past | comments | ask | show | jobs | submit | struct's commentslogin

First, I'm sorry you're having problems finding a job -- that sucks.

Second: consider that sometimes, the cost-benefit of automation depends on perspective. An example that I like to give is Ocado's automated grocery warehouses in the UK: impressive technology, very efficient, but during the COVID-19 pandemic - when everybody wanted online groceries - Ocado had to stop accepting new customers. They didn't have the capacity, and adding a new warehouse took years. The regular supermarkets hired people and bought vans, they were able to scale up.

Automation is great, but it can't help businesses adapt to novel situations. Corporate life is about cycles: the pendulum swings one way, then the other - we've just swung hard over to the automation side for now. The best strategy: know the limits of AI tools, prove your agility and ability to do the things the tools cannot do.


Or perhaps: Efficiency tends to be inversely-correlated to flexibility. Not just for companies, but but also in the natural world of living creatures.

It's a complex gamble on how the environment will (or won't) change. Both are important... but "efficiency" is way easier to measure/market in a spreadsheet.


Good suggestions, but you should also have some formal training on how to manage people on that list - every employer should offer it, and if they don't that's an _enormous_ flag.


I agree. Formal training and a manager program should happen before you get promoted. After promotion, 1:1 meetings with upper C-Level executives and/or manager tandems should exist.


It's a shame that Gemini Ultra is not out yet, it seems like a solid improvement on GPT-4. I wonder how it'll compare against GPT-5?


You almost certainly don't want to handle this at your scale. KYC involves handling a lot of personally identifying information and there are big consequences if it goes wrong. Have you considered using something like Stripe's Connect API? [1]

[1] https://stripe.com/docs/connect/identity-verification?locale...


I'm going to try and learn Rust (started and failed several times already), probably by building something with actix-web


You may also want to check out https://github.com/tokio-rs/axum if actix-web doesn't click for you.


I tried both, I'm not sure how Axum is much different than actix-web, seems like they both have similar syntax for creating a new router, adding routes and middleware, and starting the server. I think actix-web has macro-defined routes but I haven't been using those anyway.


My goal was to explore and learn Rust and/or Go in 2022... Not sure the remaining 2 weeks will cut it, so I'll have to carry the goal into 2023.


You can learn Go in 2 weeks. That's one of it's selling points!

Just dig into the official resources [1].

https://go.dev/doc/


I have a Darden-signed copy - one of my proudest possessions!


It's great that the web is so durable and long-lived, but I wonder about the health of it - it's got so complicated that we're down to only three implementations (Firefox, Chromium, WebKit), no realistic possibility of a new engine emerging, and essentially one implementation defining the standard. I wonder where we'll be in another 30 years?


This is the greatest crime against the web, IMO.

The growing complexity for the past decade has been driven almost entirely by Google. I'm now pretty convinced they did it as a part of an explicit strategy.

It's so insidious - on the one hand they are improving the web, on the other hand, the complexity they are driving makes the web more vulnerable.


> they did it as a part of an explicit strategy

Yep. First they embraced the browser by building some nice shiny, er, Chrome on top of WebKit.

Then, they proceeded to extend it, and extend it, and extend it. While paying Mozilla to stay in business, to avoid antitrust attention.

Now they've very nearly extinguished the other engines.

What can I say, they learned from the best.


It's always like this until the next new thing. The big guys will control everything until something new comes around that they don't want to implement. It will be a bit harder because you have the Web Browser which is a lot flexible than the old AOL Clients. Also WebKit is available publicly for any one to fork and create a new service.

When modems were available for residential use, BBSes were the gateway and slowly were killed by the big guys (AOL, Compuserve and Prodigy) but what these big guys refused to do is work together to allow further communications with people outside of their networks.

It took kids coming out of college and wanting to keep their internet access for email, ftp, talk, usenet, gopher and http. They started to partner with universities and offer TCP/IP (over PPP) access for $20. Local BBS started to open gateways to allow it's users to send/receive SMTP emails. By the time the big guys realized they were at a disadvantage, they started to offer communications between AOL, Compuserve and Prodigy, At first charging their users extra fees, some plans made it like SMS and were charging per message. Eventually users where just using their clients to get to the internet and that service dies. Most of those small internet providers were purchased by bigger companies.

What will need to change now to kick the big guys (Verizon, Google, Apple, etc...) in the balls again? The tech is so regulated that I doubt we will see anything new as far and networking. In NYC WiMax was hobbled by the communication companies like Verizon TimeWarner and RCN.


Yeah, this is why hosting user content scares me. AFAIK:

1. Yes - Google can safe-site ban the domain for hosting malicious content and it's difficult to get it unlisted once it happens.

2. Yes - it's always better to use a secondary domain, it just protects your brand.

3. There are cases where it's OK - e.g. when the majority of the site's content is private or substantially populated by an application (and therefore secured), or where you have a business or contractual relationship with the person hosting the content (and therefore can revoke their account on abuse).


What does Microsoft get out of this? They already have TTS and deep learning transcription, what technical capabilities does Nuance have that they don't have already (or can't develop for substantially less than $20B?)


Probably a crapton of patents for voice recognition.

Also, if you cannot operate a keyboard and must communicate by speech to operate a computer, it's pretty much Dragon NaturallySpeaking or GTFO. Integrating NaturallySpeaking tech into Windows would be a huge boon and further cement Windows as the os to have if you have disabilities.


I have users who have intentionally switched their speech engine from the latest version of Dragon to Talon, for both dictation and commands. Talon is cross platform and directly targets accessibility use cases (far more than just speech input).


I switched from dragon to talon a while back... and then back to dragon :-\ Not to bash though. You've built a great product!


I'm specifically talking about the new Conformer model, available in early access as of ten days ago. What you tried was likely the previous (circa 2018) model, which is much less accurate than Conformer.

This is a demo of Conformer in Talon: https://twitter.com/lunixbochs/status/1378159234861264896


Oh cool! Yeah I was using whatever the beta model was in December or so.


Is it slowly going to turn towards like the OS in movie 'Her', how much of it is really possible?


Nuance are absolutely miles ahead of the competition the second you're looking any other language than English.


I don't think this is accurate (and I work in this field).


And what do you suggest is better? I've worked with nearly every tool (open source and closed) under the sun in medical, industrial, and personal settings and Dragon NaturallySpeaking/Professional was by far the best in terms of accuracy regardless of prosody, accent, background noise, technical terms used, etc.

Personally I think they should've been acquired a decade ago.


That answer depends on the language and on your use case. It seems like you're asking about desktop apps, but my parent was not talking in that context. Indeed there's not a lot of choice there because there's no money in it.


I'm even talking vs custom trained models with Kaldi (was working on a startup that was trying to create lessons for public speaking so we could grab enough data to tackle accent remediation/help those with aphasic speech disorders) and again just reiterating, the out of the box performance of Nuance's products are just better than anything else.

Obviously Nuance is more than just speech recognition, but still not sure why people are downplaying how good they were at it.

EDIT: or maybe it's just too prohibitively expensive for people outside of medical/legal fields to know about? And don't get me wrong, I love that things like Talon Voice are widely available for hands free coding, I just hope this means NaturallySpeaking will supplant Windows Dictation.


If you have the data and a specific domain you can focus on then building a custom model [with kaldi] should always win. That's what I've done in the past (beating google, nuance etc.). You most likely didn't have the data and/or didn't know kaldi well.

> Obviously Nuance is more than just speech recognition, but still not sure why people are downplaying how good they were at it.

Because nuance wasn't very good.. at least in all the benchmarks I've seen. It's been a while since I compared numbers it's possible they've improved a lot. They're also known for kinda being dicks with the contracts they offer in B2B.


I've used it in medical in a multi-lingual setting and there it's basically the only game in town.


Where do you think dragon is used to its maximum efficiency as compared to siri,alexa in places of home automation and other menial tasks


Nuance owns a ton of patents and are extremely litigious.


I’ve also found the Microsoft API guidelines helpful (especially for error handling).

https://github.com/microsoft/api-guidelines/blob/vNext/Guide...


Nice reference, but like the OP article, it really goes light on security. I feel that's the hardest part of getting an API released.


Way better then this one, however only covers CRUD.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: