Congrats on the launch! Worked on something similar, so rooting for ya. Using Hollywood accent coaches is a great approach, but I'd also look into identifying speech therapists that specialize with helping adults in target regions to 1) get more information for your models and 2) point out which problems can't be corrected in older adults, so you focus on those that can.
The problem with adapting this to other forms of atypical speech is that their recommendation system likely relies on a catalog the phonemes L2 speakers have issues with (the example most people know is the Japanese "L" "R" swap) so it's much easier to create courses with specific focuses and solutions.
If Google's Project Euphonia [0] is actually still ongoing and they release their dataset/methodology of training models with that sparse dataset I can see your idea as approachable; even accented speech is a tough problem to work on considering how many variants exist worldwide (but their approach looks good!).
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.
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.
I never said zero stakes? There are clear instances where a a 1 or 0 shot transformer can have benefits beyond entertainment--topic modeling and named entity recognition for instance (I'm on the team that believes that human-in-the-loop systems will always outperform solo systems on their own and that GPT-3 alone does not confer any competitive advantage). If you think that chatbots are the only user facing use-case for a transformer, then frankly that's on you falling for the hype surrounding its language generation performance.
OpenAI knows GPT-3 is not sophisticated enough to perform medical diagnosis or analysis (anyone can look at how Watson failed), so it'd never approve such a risky application.
I mean the OpenAI team would never approve this application for production. It's very clearly stated (in both the article and use case guidelines) that medical diagnosis would be a "high stakes domain" and is unsupported. Frankly, I'm not sure why this result is even notable.
Abstracts are important (and clearly key in generating these TLDRs), but when it comes to ranking and recommending other papers (not to mention noting whether a new paper has content that can actually push a field forward) an abstract just isn't enough.
The Koch's political funding goes far beyond just this one PAC fyi. Their funneling of millions into the Federalist Society and its individual causes serving as just one example that has had an outsized negative influence on the health of American democracy.
Not only that, but the assumptions are poor to begin with. Most skincare products containing sodium hyaluronate are either micronized or in a crosspolymer so that it can penetrate the skin, not HA on its own (something I'd expect a chemist/dermatologist to know). Cursory googling it seems the writer/owner was attaching her name to reddit submissions of scientific articles unrelated to her to game SEO and give her credibility that it it's painfully obvious she doesn't have.
No, iOS devices can also soft-brick, it's just a general term. Most often people encounter the term in relation to hacking/rooting hardware, like a Nintendo DS/Switch for instance. Hitting up Wikipedia or Trends can help you dig through its usage considering how Google's search algo heavily biases current events wrt relevance.
You're going to need to provide a source for your claims that 0% graduate. A handful of my classmates from Carnegie Mellon were veterans (some with trauma) and all graduated within the typical timeline. The NVEST report from 2017 on student veterans [0][1] show an approximate 54% graduation rate within 6 years (18% were still enrolled, 28% dropped out) greater than the general averages. While the washout rate for 2 year programs is particularly high,
your idea that retraining isn't achievable is flawed.
If anyone wants to try out a free GPU using Google Colab/Jupyter (K80, you might run into ram allocation issues if you're not one of the lucky users who get to use the full amount) here's a quick guide to get a Julia kernel up and running: https://discourse.julialang.org/t/julia-on-google-colab-free...
Glad they've finally gotten to 1.0, been using it on and off since .3 with the same optimism I had when Rails jumped onto the scene. I really hope the community can start to mature beyond the bikeshedding that is typical of early language days, and start cracking more significant issues necessary for wider acceptance. Speed and multiple dispatch are great, but ability to roll standalone binaries (just don't mention it on Discourse/Gitter or risk Karpinski's wrath), streamlining documentation/onboarding, and more native/active package development are pretty big todos to tackle.
I've been working on updating a wide variety of engineering software apis/wrappers and a large informatics library for .7/1.0, but I'd love to hear any suggestions of must-need technical packages that users want from MATLAB, Python, or R.
No it doesn't. The requirements are Android 5.0, and can support something as weak as the Motorola E4 Plus[0]. And as someone who maintains a decently popular app on that platform, you'd be surprised how many former flagships (the Google Pixel in particular) still remain on Android 7.x (my wild guess is so that they can keep the old Google Now launcher) which does not have the selective permissions that come with Oreo.
Most modern phones are not on 8. The Oreo install base is only 12.1%, most phones are still running Nougat (7.0 dominates with 21.2%, then 7.1) or Marshmellow (23.5%). Seeing as PUBG Mobile's lowest supported version is 5.0, it wouldn't be a stretch to assume Fornite could do the same.
A lot of Android devices are about to have a real security flaw introduced (and normalized considering the median age of the Fortnite audience) which Epic is handwaving away.
Sorry to dump on your aside, but the media arm isn't affected. The Finance/Risk news (for financial clients) division is the one being downsized, and they were profitable.
It is indeed, correct. The annealing example I linked is from his team (here's another one with full citation with Martinis at the end there [0]). And if you want further proof that Martinis and the Google Quantum team are pursing quantum annealing look no further than the Adiabtic Quantum Computing Conference that was held this summer where their team held several talks including "Building Quantum Annealer v2.0" [1].
And I didn't touch on your earlier (erroneous) comment on the scientific community's perception of D-Wave, but I think it needs to be said that in actual professional circles the research they're performing isn't met with as much derision as they seem to garner in these more causally informed settings. It's hyped and a difficult subject to understand, so that's fair, but I suggest really informing yourself if you're going to go out there and make the claims you're making.
The UCSB professor you're referring to is John Martinis [0] and though the physical structure differs (they're using small arrays of "Xmons" instead of typical transmons qubits to minimize decoherence and thus error) they're still working within the realm of adiabatic quantum computing/quantum annealing, the same general avenue D-Wave is pursuing (with a "digital" twist) which is why Google invested in both of them.