daisystanton
In my ideal world we'll add icons for things like "does this need to be a proposition" (i.e. bond-funded measures; some pension props), a visual indicator of the $$ at stake (state/municipal assessment if necessary), progress bars for how much campaign money was spent on each prop/candidate (a nice proxy for how contentious something is).
What other metadata do you think could be useful?
Yeah, echoing another comment: we tried so many rounds of tallies/counts, because visual summaries are so useful from a UX perspective. But there are so many things that quickly start looking partisan. For example: we want to both a) be inclusive and let anyone share their choices as a voter guide, and b) not have visual summaries be a reflection of whatever biased population of users might be using OpenBallot at any time. (Some users gave us early feedback that they thought the visual summary was OpenBallot's recommendation of how they should vote, which we definitely didn't want.)
In general, we're being really conservative with UI atm because we don't want to put ourselves in the position of doing any editorializing that smells partisan. (We even have two sorting methods for guides, one of which is "random", because it's tricky to balance making the big-name guides that people expect discoverable vs hard-coding a sorting method, which we'd prefer not to do, since even that feels partisan).
We've got reams of detailed examples of these types of problems, so if anyone is interested, jump in on Discord! :-)
The guides that we ingested (from source) all capture exactly what content was available upon ingestion time (we don't yet have a way to monitor for changes and auto-refresh, but we're working on that). Those have a "verified" checkmark to show that they've been QA'd by us.
If your friend authored the guide, it means he hasn't (yet?) put his own reasons in. I'd definitely reach out to him and encourage him to add them, since I agree with you that that's what makes this kind of tool really great!
If you think there's some kind of bug / problem, please do join or Discord (I'll edit the post to include it) or just call us out on Twitter. We're very keen to resolve anything that doesn't seem right: https://x.com/OpenBallotApp
Thanks so much for the feedback!
Two options to contribute new guides: 1) Tweet https://x.com/OpenBallotApp with the guide URL, or 2) join our Discord and post in #guides-wishlist: https://discord.gg/mzubN2v4
We'd love to hear from you!
The site is our first prorotype; if you're in SF, please try it and share your feedback! We're hoping to use this to eventually scale to all over the country.
[1]: In 2006, the NHTSA adopted a "footprint" approach to CAFE standards, where fuel efficiency standards are set based on the vehicle's wheelbase multiplied by its average track width. I haven't dug into why this is the case, but people seem to agree that this results in less stringent fuel economy targets for larger-sized vehicles. See https://en.wikipedia.org/wiki/Corporate_average_fuel_economy
[2] A key piece of U.S. tax code, Section 179 allows businesses to deduct the full purchase price of qualifying equipment up to $1M(!) For trucks, the vehicle must have a Gross Vehicle Weight Rating (GVWR) over 6000 pounds to qualify for the Section 179 deduction. This incentivizes businesses to purchase larger, heavier trucks. See https://en.wikipedia.org/wiki/Section_179_depreciation_deduc...
[3] A friend told me (citation needed!) that the ability for a truck bed to carry 4x8-foot drywall sheets is often advertised as a selling point (is this true?). I suspect a lot of this is just the US market lusting after "bigger is better", but I'd love to hear anything others know about this.
Anyone have details they could add to this?
* The Auzzies had an old version of this: https://en.wikipedia.org/wiki/Ute_(vehicle)
* There's an awesome Tesla pickup truck mod that I'd buy long before the Cybertruck: https://www.youtube.com/watch?v=R35gWBtLCYg
Analyzing latent embedding capacity in Tacotron, Google's seq2seq TTS model
Research paper: https://arxiv.org/abs/1906.03402Audio examples: https://google.github.io/tacotron/publications/capacitron/Capacitron is the Tacotron team's most recent contribution to the world of expressive end-to-end speech synthesis (e.g., transfer and control of prosody and speaking style). Our previous Style Tokens and prosody transfer work implicitly controls reference embedding capacity...
Think you can train a model to detect Google's synthesized speech?
Predicting Expressive Speaking Style from Text in End-To-End Speech Synthesis
Expressive Speech Synthesis with Tacotron
Quantum computers are about to get real
Deep Mind Launches Deep Mind Health
The wiring diagram of the male nematode’s nervous system is only a beginning
Human-level control through deep reinforcement learning
Crowdpac.com: using data science to help inform voters at election time
https://recode.net/2014/06/02/new-approach-to-nonpartisan-voters-guide/
"Crowdpac uses data to inform its recommendations for federal and state races. It analyzes stated positions on key issues, an incumbent’s voting record and campaign contributions to score each candidate.Voters take a quick, three-step quiz to determine where they fall on the political spectrum, from liberal to conservative, and identify which issues they care most passionately about — say,...