Our designers use Figma but the licenses are so expensive that people on the product team can’t even make small changes or illustrate concepts. Additionally the UX has tons of hidden features making it hard for non experts to discover how to do simple things like export mockups.
As a result the designers have a backlog of small busy work changes instead of focusing on the key design questions. No one else has licenses to make updates or knows how to use the tool.
The product team has started using Whimsical which is easier to use and is more reasonably priced for people who only use it occasionally or need to quickly illustrate low fidelity concepts. Now we have two design tools to essentially get around UX and pricing problems.
> […] the licenses are so expensive that people on the product team can’t even make small changes or illustrate concepts
I completely agree.
I don't see the Figma files as an output that designers produce. I see it as a product on its own that people should be able to collaborate on.
I'm not going to stop a designer from sending me pull requests. In fact, for small mundane updates I'd LOVE it, since it would take this burden away from me. In fact, our inhouse designers all have GitHub accounts, so in theory they are able to.
But when I want to tweak a little thing in a Figma file, like adjust naming of the symbols, clean up icons, add more diverse examples of the content to document edge cases, or to have grounds for my next discussion with a designer or engineer, I'm not allowed to. Also quite often the thing that I want to do is not necessarily meant for designers, but for my fellow developers. Things like technical notes for implementation.
> the licenses are so expensive that people on the product team can’t even make small changes or illustrate concepts
Licensing models that don’t scale can cause tremendous problems. Early on, design can become a bottleneck. Later, simple changes might be rejected because the cost of redesign is too high. In the worst cases, developers will work ahead of design to meet deadlines (since not every team can afford an expert in the design tool) and the resulting variances will be challenging to reconcile/resolve.
Expensive, high learning curve tools can introduce silos and bottlenecks into an organization.
I love Whimsical. It's just so damn smooth and reliable.
All that being said, I think wanting one tool to serve not just designers but also product teams is probably too high of an expectation. We don't blame developer tooling if designers have a hard time making small updates to demonstrate design concepts. They're different domains, so different tooling is expected.
It amazes me that there is no authentication provided by governments in the US to citizens. They just accept a social security number as if it was some sort of password, when it was never intended for that purpose. Other countries give citizens an electronic ID to authenticate themselves. It seems this would prevent hundreds of millions of dollars in fraud and identity theft.
UI is a program geared toward optimizing payment speed, with fraud looked at as something that is addressed via audit. Historically most fraud or bad reporting can be addressed by capturing future benefits.
With the unprecedented load being placed in these systems, you’re going to see things like email and sms used more, which enables new paths for fraud. Pandemic unemployment is also geared towards gig economy workers, which again is a new frontier of fraud.
It removes a lot of friction if you police transactions after the fact. The only trick is to make transactions un-doable if they turn out to be fraudulent.
What we got instead is Real ID, https://en.wikipedia.org/wiki/Real_ID_Act?wprov=sfti1, which is a set of guidelines that States and federal agencies must follow to authenticate people for the issue of their ID and anti-counterfeit features that the ID should have. In other words, the issue was put onto the states.
Reason #4: ID cards would function as "internal passports" that monitor citizens' movements
I don't understand this one. This was never a thing in the EU, even though IDs are mandatory in just about every member state.
I spent on-and-off four years in Italy and while I initially had to present and ID to my landlord there, who then needed to pass this data to the police, nobody bothered me after that or checked if I'm still there.
Hell, even after a law was passed that initially basically forbade anyone who was in the country more than half of the year from driving a car with foreign plates I still wasn't bothered by anyone, because as I was a citizen of a Schengen area state, there was no reliable way to determine when and where was I lately.
That's interesting. When I changed jobs and moved from NY to Indiana, so my wife could pursue a graduate degree, I had every intention of maintaining my ID and permanent residence in NY, (since I could always still receive mail there via my parents, who allowed me to maintain my permanent residence there whenever I rented or was resident in student housing.)
It quickly came to my attention by communicating with car insurance that I could not do this legally (they sought me out, I have no idea what caused this, perhaps a National Change of Address record triggered?) my car insurance would be terminated because my car was no longer "garaged" in NY, and a lack of insurance on my vehicle registered in NY would trigger a suspension of my license, (and eventually a bench warrant could be issued potentially leading to my arrest, if I did not take action before 30-60 day window passed.)
I wonder if you got lucky, or if this scenario doesn't play out the same way in EU? FWIW, it turned out that everything about being an Indiana state resident is cheaper than living in New York, and it really was to my benefit to get my home permanent residence changed to the new state.
(It was very surprising that I had to do this, though, as a student you are allowed to maintain your primary residence in a different state, I guess this justification works for undergraduate but not for a spouse's PhD study...)
Car insurance and registration is one of those "interesting" areas if one bothers to peek below the surface. I've got a couple stories about it, but how about this (details removed to avoid personal information) one.
A few years ago, my girlfriend moved overseas for about a year, nearing the end of her time overseas I went over and we got married as we had planned. A short while later she returned and moved in with me, having mostly gotten rid of her car/apartment rental/etc (and moved the remainder of her personal items she didn't take overseas to my place) before she left the US. Within a couple weeks of her return, I received a letter in the mail from my automobile insurance stating that they had reason to believe that additional adults of driving age and related to me were living in my house but weren't on my insurance. I either had to notify them of said persons and sign some paperwork indicating that they would never drive my vehicle, or I had to add them to my insurance (for an additional $$$ a year of course).
Now when I got married overseas we did some some paperwork local to that country. But the state I was living in, there was additional paperwork that needed to be completed stating that I had been married overseas/etc. As far as I'm aware that paperwork had not yet been filed before the insurance company contacted me. Nor had my wife changed her address from her foreign one.
So, somehow, not only did the insurance company discover that we were married, they somehow found out when my wife had flown back to the US as well (she returned a bit after me for various reasons). Its not hard to come up with ideas for how they might have put these details together, but I've never managed to find any evidence of the existence of the kind of channels/databases that must have existed for them to pull this off, considering it was a low key event.
In Oct 2020 a passport or state-issued Enhanced ID [0] will be required to board a domestic flight in USA. It's about as close as they could get since no one wants a "national ID card".
How about making the id voluntary to get, but required to get benefits. Want to get the guvmint out of your life? Sure, then don't ask for unemployment benefits.
As an American: see the classic "Get your government hands off my Medicare" line. I don't know how many of us actually paid attention in Civics or bothered actually trying to understand how our government is supposed to work.
How about you get evicted and cannot get food stamps for your family, because your non-driver ID expired three months ago and you have a hard time getting a day off to take a bus downtown to DMV?
There are clearly other ways to solve this that don't involve depriving people of food stamps. Most every other developed country has figured out some solution.
First of all, in most countries an expired proof of citizenship is accepted for many purposes because it's assumed that people didn't go out of their way to coincidentally lose their citizenship or permanent residence when the ID expired. If it proves residence or driving qualifications, then there are certainly other reasons why it should expire.
Suppose we have an administration, decentralized or otherwise, that stores the records of the people concerned. They can then be contacted and details can be verified.
This informal verification already occurs on many levels, particularly in the US due to the lack of consistent ID. Try flying on a flight without photo ID, entering the US as a US citizen without proof of citizenship, etc. You will be permitted to do so with a bit of extra hassle while you're identified to a reasonable degree of confidence.
I've found that HN and other online communities have a disproportionate number of users who have no idea of rural life in America. As such they cannot fathom a poor, rural person without a birth certificate or a photo ID or the ability to get either.
edit: To get an EBT card in NYC you can do it all online if you have a valid (ie, not expired, ID card.) If you do not have a valid (ie, expired, ID card), then you have to go to the DMV so they can take your picture and you sign a few forms. The forms are available in 22 languages. At the same time they may work to get you a new, valid, ID card.
How are the DMV opening hours/wait times? Now imagine that impact on a person with some minimum wage job. You have a valid point but the marginal cost of bureaucracy to a poor/disabled person is often a lot higher than to someone for whom life is going smoothly. Also, it's easier to fall off the smooth track than to get back on.
What is the alternative to get a valid ID card and an EBT card? I understand there's a hardship for someone that can't get away for a few hours to travel to the DMV office. But it's the same hardship for everyone. There are basic requirements:
1) You have to go to the DMV office
2) You have to agree to have your picture taken
3) You have to fill out 3 forms (offered in your native
language -- 22 languages are offered)
4) You have to provide a mailing address for where the EBT card will be sent
And then you have to be on the other end of that mailing address to receive and activate your EBT card.
This all seems like very easy procedures to follow to get food stamps.
I understand there's a hardship for someone that can't get away for a few hours to travel to the DMV office. But it's the same hardship for everyone.
No it isn't. The marginal costs are different. If you earn $2000 a week and through some mischance have to give up a day's earnings to go the DMV your $400 loss is an annoyance. If you earn $500/week your loss as a percentage of income is the same but the economic impact of losing $100 is probably much bigger.
> "Voluntary to get, required for some benefits" is another way to say "involuntary". What is citizenship but a collection of benefits?
This is pretty clearly a poor extrapolation. For example, Global Entry. Is signing up for Global Entry involuntary? It is voluntary to get, required for some benefits.
You're not getting any intrinsic benefits. If you are a US citizen, you are allowed to return to the US after international travel. Global Entry doesn't change any of that.
On average, it does make returning easier, which is nice... but the machines could be out of order, or you could be flagged for questioning in the usual manner, etc.
If “making something easier” doesn’t count as a “benefit”, I think that maybe there’s a fundamental disagreement about what it means for something to be beneficial.
Global Entry is pretty clearly beneficial for the user, as part of the border control experience. Whether having things like Global Entry is beneficial to society is, as `tptacek points on parallel to your comment, a very different question.
The main purpose of government is providing infrastructure like roads and bridges, as well as enforcement of property rights and security through police and courts, as well as through healthcare and armed forces.
You get all of that without this hyothetical ID. Unemployment benefits is somewhere much further down the list. It could be argued to be a security measure both to keep the crime rate lower and to prevent an uprising from disenfranchised poor people, but it serves this purpose just fine even if a few people voluntarily opt out.
The main purpose of government is providing infrastructure like roads and bridges
Roads and bridges being a government function is a somewhat recent notion that we've grown accustomed to.
Historically in the United States, roads and bridges were privately owned, and users paid a toll to a private person or company to use them. This was one of the many disagreements between the states that led to the Civil War.
There are plenty of private roads and bridges still in existence in the Untied States, mostly in the older states.
> The main purpose of government is providing infrastructure like roads and bridges
> Roads and bridges being a government function is a somewhat recent notion that we've grown accustomed to.
> Historically in the United States, roads and bridges were privately owned, and users paid a toll to a private person or company to use them. This was one of the many disagreements between the states that led to the Civil War.
> There are plenty of private roads and bridges still in existence in the Untied States, mostly in the older states.
I've always wondered about the bridge at Dingman's Ferry. Reading through the website, I wonder how they could possibly enforce the penalty for overages in terms of tonnage. Since they are a private entity would law enforcement issue a citation or would the bridge corporation be forced to litigate?
Wouldn’t anyone who did that be subject to a civil suit for damages? Also the bridge owners would have insurance.
In addition to standard economic devices like tort and insurance, the bridge owners could have a part of the road before the bridge that is designed to buckle or alarm if a weight is exceeded. That would save them a lot of money and frustration.
I figured as much with regards to civil suit, was just curious about public enforcement of private regulations when the lines appear blurred. Further, I wonder by what authority they can even set monetary fines? Like, why stop at $X for a fine? I ask because their site lists specific penalties which seem somewhat arbitrary [0]. I can't arbitrarily "fine" someone $1000 for stepping on my lawn. I can certainly take them to court for trespassing and possibly collect some damages, but those damages are not a fixed value in a fee schedule. So I wonder how this corporation has the authority to impose fines.
It looks like it's not a fine in the sense that refusal to pay can result in suspension of your driving license and possible wage garnishment. They don't even call it a fine but a "penalty". Basically they ask you for $50 or $100 depending on which limit you exceed, and refusal to pay risks a court case. I'm guessing the bridge needs to be inspected after the weight limit is exceeded or if a taller vehicle strikes the structure. The cost of inspection likely exceeds the penalty. They could easily ask for thousands of dollars in compensation. And even if you win the case, you have to pay for a lawyer and spend time in court. It's easier for both parties if the driver just pays the penalty.
I don't think it's a particularly recent notion. Ancient cities are the archetypical government, providing defense, some sort of justice system, and (often paved) roads. We see evidence of that from as far ago as the nearly 6000 year old city of Ur. Where larger empires existed, they often built larger road networks between cities to facilitate commerce and troop movements. The Inca road system and the Roman roads are well known examples of road networks built by their respecive empires. The Romans are also kind of famous for their bridges (viaducts and aquaducts).
Of course the less important roads were and still are often private, and the early US had an atypical lack of government that made this more common. But I don't think that proves that governments providing roads and bridges is a recent phenomenon, it's in fact rather ancient.
The US government exists to collect taxes, pay debts, prove for common defense and provide for the general welfare.
Roads were historically a local and state priority, so be careful with your modern conservative principles, as they probably are not compatible with your lifestyle.
How is that different from requiring vaccines for public school, a license to drive a car or fly a plane, or even a safety course and hunting license to hunt? Most of the things that are benefits of citizenship that don’t require any voluntary steps are true “public goods”, like national defense or the societal benefits of education, etc.
It seems more like the whole "Raise your drinking age to 21, or the federal government will withhold road improvement money from your state." It's coercion.
Employment is also voluntary yet we are somehow okay with it being necessary to not become homeless and will often accept terms which are very biased towards the benefit of our employer.
We have those. They're called Passports. But the rub is that some states and local municipalities will not accept a US Passport as ID. Which makes no sense what-so-ever.
This line of thinking sounds consistent but actually isn’t: even if you’re against the government interfering in your life, you’re still entitled to the benefits that you paid for. Your line would only be consistent if the individuals could opt out of paying. This is the source of the “coercion” claim that libertarians make.
First of all, an ID doesn't need to have anything to do with citizenship. It can also be a claim of residence like a driver's license in the US.
Second, if people are eligible for benefits, they are clearly being recorded in some fashion. If the benefit requires permanent residence in the US, I would presume most states are attempting to verify this as well.
In either case, this can be used for either a residence ID or a stronger ID that proves citizenship or immigration status, the latter resembling the national ID cards that many EU countries (among other places) have.
USA is set to require a passport or state-issued Enhanced ID [0] for domestic airline travel this year. It's about as close to a "national ID card" as it can get.
>First of all, an ID doesn't need to have anything to do with citizenship
Here is the comment further up that this comment is in the context of:
>It amazes me that there is no authentication provided by governments in the US to citizens.
Do you see why we are talking about citizenship now? Especially when much of the discussion is revolving around voting as well, which does require a certain citizenship status.
The discussion was about unemployment benefits, which in most every case is not limited to US citizens. I believe that when the person who you quoted used the term "citizens," they meant it in a looser sense to refer to people eligible for unemployment benefits, which is what I responded to. Citizenship is also not sufficient proof to receive benefits, so I'm unclear why we're trying to add another confounding factor when states already have a (less than comprehensive) system for tracking residency that can be adapted.
You can get an ID without a birth certificate, many people do. And it has nothing to do with minorities; a large percentage of people without birth certificates are white.
Or social security numbers. Or tax returns. Or proof of address. Or... basically anything that can reasonably indicate that they are who they say they are.
Voting is a much more important right than the other rights, because voting is fundamental to the existence of a republic. One could argue that the right to bear arms exists for the primary purpose of protecting the right to vote.
Voter ID fraud is exactly the kind of thing that infringes on right to vote. Stronger protections on voting is what protects this right, not the other way around.
As someone who not only lives in a country with a widespread voting fraud (done by government officials), but also have been an observer on number of elections and have seen this taking place first-hand, I can't understand how relaxed are Americans about this issue.
> Voter ID fraud is exactly the kind of thing that infringes on right to vote.
There's a difference between "infringing on the right to vote," which is where you're literally preventing from someone from voting, and "diluting a legitimate vote", which is where your vote doesn't weigh what it ought to. Mathematically, it's the difference between scoring a zero and scoring some fraction less than one.
It turns out that, at least in the USA, advocates of voter ID requirements and other unnecessary impediments to voting in fact desire the opposite effect - that their votes be worth more than they would be if widespread voting by qualified citizens were easier than it is.
> I can't understand how relaxed are Americans about this issue
We're relaxed about it because the data (and we have measured and investigated, many times) says that voter fraud here is so rare that it falls well beneath the noise floor of statistical significance.
The bigger issue is that people who need services like unemployment and food stamps most have low penetration rates for things like valid government IDs.
That's not really true. The Right is generally united on the point of wanting universal ID. Nothing totalitarian about a nation being able to reliably identify and distinguish its citizens.
Unfortunately the political Left believes that such ID, specifically when used as a means of election security, would lead to discrimination.
You are conflating two things: a national/"universal" form of ID, and voter ID.
Voter ID is the requirement to show ID at polling stations in order to vote. That's what the left is generally concerned about. It's a separate concern from whether a national ID card ought to exist.
On the other hand, the existence of a national ID card is generally opposed by people on the right, which is the opposite of how they feel about voter ID.
If it were assigned for free when you were born and there was no effort associated with getting it or working with it, then there would be no issue. The current problem is that a driver’s license takes a long time to obtain (because the DMV wait time sucks as we all know), and because it’s not free. This means that it’s a lot harder for someone holding down 3 jobs or working during DMV hours to get one. You are basically making it more difficult for an already under-represented group of people to vote. It’s not that it’s impossibly hard or totally preventative, it’s just another obstacle.
Those are both pretty simple issues to address. Many Democracies around the world use some form of voter ID and we could easily just follow their implementations with some adjustments.
If you are a citizen of the united states, you get a vote if you're 18, according to the constitution. No tests, IDs, or other things are required. To add any additional burden is counter to the constitution, and as a result any additional burden could be seen to prevent people from voting that have the right to vote.
Nevermind that when you add additional barriers, discrimination occurs against anyone that cannot meet the barrier, or does not want to meet the barrier.
Example:
- "tests" in the South during civil rights to prevent african americans from voting
- Requiring any sort of payment or money to create a Voter ID in a state. If the person does not have money or time this is discrimination and against their rights as citizens (you are not required to prove you are a citizen. your ballot can be provisional)
- Requiring someone be able to read. It's not a requirement to vote. Any forms requiring reading are a no-go.
- Requiring them to have a permanent address (again, leads to discrimination for those without addresses.
- Requiring someone take a lot of time they cannot afford to get an ID (again, some folks are working too many jobs to go to the DMV for a day)
Some places have tried to institute voter id laws that require ids that are difficult/expensive/time consuming to get, sometimes specifically making it harder for the most downtrodden segments of society to vote. That's really bad and so there's an outcry. Sometimes the nuance of "discriminatory ID requirements are bad" gets lost in the zeitgeist and circulates as "ID requirements are discriminatory and bad."
You might just be surprised at how many US citizens do not have a state issued ID card. There are just a lot of poor people who can't afford to pay for the ID or their parents never kept their birth certificate and they just don't have the slightest clue what to do to get another birth certificate. It perplexes me, but some people are just that broke or just can't get it together enough.
Maybe the commenter above is thinking of a different point, or coming at it from an oddly phrased perspective.
The Democratic party relies on a certain segment of immigrant or immigrant-related citizens to vote in support of them. And if licensing / IDs are perceived to target and identify who is not a citizen (your relatives, friends), then they could lose support. I suppose it could be seen as a kind of "discrimination". And if some social services, policing, etc were to be able to use such ID, then illegal aliens would certainly be more at risk of being discovered or face more stringent (less porous) treatment in the law enforcement system.
I personally think this is a ridiculous situation from every angle, and unfortunately it's all tied up in our immigration and economic policies, so it's hard to disentangle or fix.
Give the Id to everybody who wants it for free. If someone cannot prove citizenship, but they can prove having worked or lived in the US for more than 5 years (checks, bank receipts, etc.), give them citizenship.
There, problem solved. That way, you only discriminate against those who are either in the US illegally and are not working, or are working but have been illegally living in the US for less than 5 years, and both situations are fixable by the individuals themselves (work for 5 years and "earn" your citizenship).
Of course, what many want is an Id that can actually be used to prevent poor people from voting, while also being able to employ those same poor people at very low rates using the fear of "reporting them".
This is quite a common claim from the Left in the US. If you do a quick search in Left leaning publications on the issue of voter ID you'll see that stance dominates their discussion of the issue. I agree it's an extraordinary claim, but it's a conspiracy they've latched onto.
You can in fact see a child comment below where someone is commenting that the purpose of such ID is to disenfranchise the poor.
> This is quite a common claim from the Left in the US.
What about the source? Are you able to find anything that corroborates your extraordinary claim? Because I asked for a source, and you just reiterated your baseless assertion.
I dunno, this claim isn't extraordinary. MSNBC/CNN/Nytimes/WashPost say stuff like this all the time. It's a very common talking point. Certainly (in my opninion), the reason why some people are so interested in voter ID is to make voting harder.
> Voter ID's are discriminatory or pushed with discriminatory intent:
Your own link does not support your baseless assertion. The only claim is that so far voter ID laws have been crafted to exclude non-white US citizens from the electoral process.
Taken from your article:
> New studies suggest that the motivation of these laws is suppressing non-white voters, and worryingly, that they will be successful at doing so.
Do notice that the remarks refer to voters (thus, citizens with the right to vote) who, due to their race, are being excluded from casting their vote.
If that's the best source you managed to produce then I'm afraid that you were either lying or very confused, because your original claim has zero basis.
I am as conceptually socialist and communitarian as they come but there are many reasons for the federal government not to be the identity provider.
It should set the rules (the GOVERNance) by which identity providers provide that service, but it should not itself be in that business.
My favorite way of thinking about it is- the US federal government is a singleton. In any system, you want your singletons to operationalize as little as possible, because they are hardest to change.
Another way- the US federal government is an immortal entity. It represents a perpetual accumulation of all kinds of debt- legal, administrative, technical, financial, whatever. Building and scaling new operational systems within an infrastructure consumed by debt is doomed.
The thing it can do is creating the rules and policies by which a federation of private entities can operationalize a particular need. These entities have limited lifespans, can fail, and have profit and efficiency motives, can compete for business, and are overseen and supervised.
This structure exists in lots of areas, and is more successful in some- banking- less in others- military contracting. But it's vastly preferable to that work being done in the singleton itself.
If governments are singletons, what are individuals? Maybe individuals are other objects? And now the individuals don't need to hold a reference to a government service object if they want to authenticate a message they get from another object. They just ask an identity provider object. But which one? Do individuals have a list of identity provider objects? But what if the sender is using another identity provider object they don't have? Ah! The message could contain a reference to an identity provider. But why should the receiving object trust it? Wouldn't it have to ask a government service whether the identity provider is to be trusted? No silly, we don't want a government reference! It could ask other individual objects whether they trust that identity provider object! Then cache the response? Help me out here, how does reputation work?
Seriously though, you're just moving the problem around. Adding complexity. I mean, does an identity provider object still respond to messages when it's entered bankruptcy proceedings? If you're going to use an analogy, find one that informs.
Snark aside, agree that identity is a complex problem- and compartmentalizing it into components with well-known lifecycles that have known failure modes is the right solution.
The alternative- a single monolithic identity system? No, thanks.
Note- large governmental IT systems underlying programs like Medicare and Medicaid are not operated by government employees, they are operated on a contractual basis by large IT shops. You just don't know who the operator is. That's arguably suboptimal- but a different conversation.
To the specific question- what happens in this model when an identity provider goes into bankruptcy- the same thing that happens when any entity providing critical services goes into bankruptcy.
When a consumer-facing bank fails (for instance), the bank's customers
a) don't lose their money
b) don't lose access to banking services
Their accounts are taken over by a comparable entity operating in the same geographical area.
When a critical insurance provider fails, the other entities providing comparable insurance in the operating areas have to take those contracts (even if they are terrible contracts, which they likely are, because they caused the provider to fail).
It doesn't always seem like it, but this kind of market partitioning and supervision is something that in the US both federal and most states do quite well. We should have more of it.
If you're right, then you can institute devolution of any of those things and the issues you cite are going to build up over time. It will just happen at different levels in because it's spread around some many different systems.
The unemployment issue is an excellent example of this. You want to know the reason why Congress gave a flat $600 to all UI recipients even if it would be more than they were making before? Because there isn't a single unemployment system, there are 50 systems each unique, each with their own "debt", and trying to implement appropriate strictures in all of those systems would have delayed that part of the stimulus for months, if not longer.
The way you deal with those issues is by having infrastructure that is built to deal with the issue. Call it societal/social/legal "garbage collection". Whether or not we have such infrastructure, you don't get rid of the problem by shuffling it around.
I really hope you adopt the latest models particularly streaming attention variants. I think you should validate with users the assumption that latency is more important than WER.
IMHO the WER is more important than latency improvements in the millisecond range. The most frustrating thing is having to dictate over and over and the transcription is incorrect each time.
Consider that the time to a correct transcription is the latency plus error correction. If error correction is manual it will be orders of magnitude slower, so optimize for WER.
I’m terms of competition, Siri has latency in the 5+ second range due to the network call especially in area with poor data rates. I think a client side model like yours will easily win in this category. If you’re already ahead here, why not focus on WER next?
Another great capability is to generate alternative transcriptions for words with low confidence values to allow for quick error correction. Do you offer something like this today?
Also, consider the long term view that new models are constantly being released and refined. It’d be best to have an architecture that allows quick replacement without a lot of hand tuning, or where the tuning can be automated to a greater extent.
Thanks for all the hard work you have put in so far @reubenmorais
+1000 to @mostlyjason's comment - Great latency figures mean nothing if the word error rate is high, since it dents confidence in the output (so why use DeepSpeech?) and (as the parent comment notes) necessitates manual error correction.
I would love to see a future release focus on optimizing WER for these reasons.
If PayPal wasn’t such a crap company they’d be the perfect solution for this problem. Just enter your payment and shipping info once and get one click checkout on any site.
Just skimming through the article you posted, it says trials with a similar treatment resulted in one patient dying from a massive immune response and two other got cancer. Sounds like it relies on a virus to deliver the DNA, but I didn't read too carefully, so maybe CRISPR will turn out to have fewer side effects or a greater success rate. I'm pretty sure one of the up-sides to CRISPR is compared to other gene-editing techniques it's relatively cheap, which is always a plus.
You can just run any old DNS server with IPs that you can only access if you also happen to be on a particular ZeroTier network. Route 53 works.
The tricky part of ZT-only DNS would probably be making it work for more than one network at once. So you couldn't send all DNS to one place that serves *.myzt.example.com and recursively resolves anything else, because it wouldn't know about myzt-2.example.com.
Edit: a Route 53 map of your network may be easier to achieve if you use the terraform provider, which I have let languish for a while and should probably give to someone else to maintain (ZT people -- want to take it on? You've got Go expertise now!): https://github.com/cormacrelf/terraform-provider-zerotier
I can confirm I have an android TV with sideloaded zerotier (would be great with official Android TV support, BTW! Ed: Android TV has a separate app store - apps need to declare themselves compatible), a few Linux boxes - and mDNS/bonjour works fine.
This is pretty cool my biggest problem with micropayments is friction when browsing and friction for managing subscriptions. Integrating it into the browser should make it much more automatic.
It’s kind of a chicken and egg problem where publishers wont benefit until readers have it and vice versa. I hope this grant results in a killer app that generates some critical mass for the standard.
As a result the designers have a backlog of small busy work changes instead of focusing on the key design questions. No one else has licenses to make updates or knows how to use the tool.
The product team has started using Whimsical which is easier to use and is more reasonably priced for people who only use it occasionally or need to quickly illustrate low fidelity concepts. Now we have two design tools to essentially get around UX and pricing problems.