Subscribe
Sign in
Golden — Mapping human knowledge with AI
The next place for canonical knowledge on the Internet that will eventually cover 10bn+ topics. Golden, among other features, has a WYSIWYG AI assisted editor that accelerates the process of extracting information, applying citations, filling tables and more.
108
Replies
Jude from Golden here. Working hard to get the number up and feel free to help us. This issue will go away over time. Thanks for testing!
@judegomila I think that we might have discussed this before, but what did you consider RE building this on the blockchain?
I could think that there could be some argument that it could help accountability about changes etc? what did you consider when making the decision not to?
@_jacksmith Yes I have discussed specifically with Juan about IPFS and other technologies that could be used to make backups onto the blockchain as we go along. It seems to make sense as a direction and I'll be looking for more experts to talk to on this.
@judegomila @_jacksmith lol, right? perhaps it's the internal name for the ai. 😂
i will elaborate more on this question later, but basically: have you or Juan considered to not only "backup/sync" to Blockchain, but going one step further and making it a sort of torrent spreadsheet/database for general knowledge things?
i'm specially interested in a kind of opensource craigslist with a map and without all the spam!
easy peasy. 😁
I've been a user of Golden since August 2017 and have contributed towards hundreds of topics.
As someone who's previously TRIED to spend time contributing to Wikipedia, I can say that they really make it hard for a layman user. The editor is confusing and can require you to use "Wikitext" markup language to achieve what you want, I also ran into bureaucracy in the way of random rules and politics, all whilst just trying to be helfpul. I'd spend hours writing a topic on Wikipedia only to have it deleted for some random reason that I didn't fully understand.
Golden addresses a lot of these concerns, it's a purely WYSIWYG editing interface with no coding needed; I was also able to submit content without much by way of random rules. That said, I did find the difference between "clusters", "topics" and "categories" to be confusing - I still don't think that I fully get that part.
The AI editor is pretty raw in terms of its functionality, but it's clear to see its early promise - it can pull in content from other websites and format it all for you etc. very cool and saves a huge amoutn of time.
As a user of the site right since its very early inception, I have sometimes come across bugs (e.g. posts getting lost in my drafts folder), but the team has addressed these very fast when I've reached out, many times within a few hours.
I look forward to continue contributing to Golden, I find it very fun in a way that I had wished Wikipedia would be when I first started trying to contribute there.
+ Don't need to deal with bureaucratic rules/politics like on Wikipedia + I like the easy to use editor + AI editor has early promise
Cons:- I don't understand how the leader-board algorithm works - Some bugs - Some parts of the site are confusing
I just saw it. For those who are not on @ProductHunt and will not read this thread, how will they understand what is Golden. I could make little sense out of the message on home page because it sounds like just another (comprehensive though but the fancy AI mention is a turn-off) curation-KB-community tool. it does not inspire enough confidence that I should spend time on it. Sorry.
Jude CEO and Founder of Golden here. Super excited to take this live. We are out to build the next place for canonical knowledge on the Internet. It has been a long-term mission for me to open up the knowledge coverage of billions of niche topics, companies, technologies and new concepts. Our aim is to cover in excess of 10bn topics in high detail over time. Although we all love Wikipedia, there have been various issues in the last 18 years, from constant deletion of data (product hunt was almost removed a few months back) to fact validation and automation of processes/work and UI/ease of user. We also believe there are many more features that users want, like a knowledge feed, keyboard commands, AI assisted feedback on editor contributions and tables that can automatically update.
We have set out to:
1. Cover all topics that exist over time rather than just ‘notable’ topics.
2. Go into greater depth around a topic, from its timeline to videos and other useful resources surrounding the topic (eg learning videos, further reading, blog posts, Q&A, podcasts etc).
3. Support a larger population of people trying to learn about topics.
4. Make knowledge more accessible, richer and fun to read about.
5. Allow you to track topics of interest and be updated when new information is available on the subject.
6. Save time making the knowledge in the first place by using design, UI and AI to aid construction of the information. Especially by automating repetitive tasks and bring smart editor features.
Initially we have kicked off with various areas from ‘cell and plant based meat’ to synthetic biology to cryptocurrency consensus mechanisms to artificial intelligence, microbiome, stem cell technology and startup topics. We expect these areas to increase in scope over time covering space, medical food, clean technology, robotics and many more exciting fields.
We are still early on our journey to delivering our vision and very much looking forward to product feedback and help with building up the content. Our team is hard at work making the product easier to use, we are up for taking every flow to its simplest form and removing every bug. If there is a feature you have been dying for on Wikipedia but could not get it, please also let us know. We look forward to seeing you in our community and covering topics especially under represented elsewhere.
You can find out more about what are making in this blog post here: https://golden.com/blog/introduc...
@imromains Good questions. The topic pages are put out on CC-4.0-BY-SA ie open for the world. We make money by charging companies to do advanced queries and use our AI tools on their own data. The more correct and neutral we are, the better our product is to both the public and the paying customers. Thus we can be impartial as there is a product and financial incentive to do so. Additionally, by working with companies we believe we will be able to get more information open sourced quicker - just like github does. Giving WP that funding won't solve the issues discussed in the blog post here https://golden.com/blog/introduc...
@judegomila @imromains Impartial to Golden atm, but most people don't know that Wikimedia, the parent company of Wikipedia, makes $100 million in annual revenue. They don't really need more money from a tiny startup.
Source: https://en.wikipedia.org/wiki/Wi...
@kirillzubovsky does Wikipedia breakdown where that $100mil is earned from? I know Google and Mozilla sometimes donate money etc
@kirillzubovsky @_jacksmith Here's their annual report, it doesn't break down large donors though (e.g. doesn't show how much from each donor). They brought in $91.2M in 2017, $81M was donations and $1.2M investment income.
https://upload.wikimedia.org/wik...
Can you tell us more about your AI strategy? The recent boom in natural language queries (primarily through voice search) creates a need for an extensive knowledge graph. If your service can have more raw data, but also more "smart data" than Wikipedia/Google's graph, you can win big times.
Also, if that's something you'd be open to discuss in person here in SF, let me know. I founded a startup doing NLP technologies for developers to easily add voice control to any code on any platform, and am a deep diver of the semantic Web.
1. The reviews all look like shills (friends of the founders). Golden basically has little to no content compares to wikipedia. (Tried a few topics, all have no articles). I don't understand how you are getting such glowing reviews on product hunt
2. This is ambitious. How are you going to make sure that this is going to turn out differently from Microsoft Encarta. Why would millions of contributors contribute to you and help you make money?
3. You have very little content. How do you plan to bootstrap your content? Do you plan to focus on a particular set of topics first?
I randomly tried a couple of things, no articles:
https://golden.com/wiki/Depth-fi...
https://golden.com/wiki/Maximum_...
https://golden.com/wiki/Ordinary...
4. I don't understand why everyone is so hyped about your UI. Sure, it looks fancier and maybe it's better for getting those VC dollars. This is pretty slow considering how little you are doing. Wikipedia is extremely fast and loads fast in basically every browser.
The UI is also definitely not more usable than Wikipedia, let's compare the 2
https://golden.com/wiki/Ethereum
https://en.wikipedia.org/wiki/Et...
The golden UI is prettier at a glance, maybe. But the wikipedia UI is just more functional. The font color of golden is softer and has less contrast. the lack of divider makes it harder to realize that you are going across sections
@sidazhang I don't know about the reviews, the ones that I read seem to be related to the functionality of the editor vs the density of content listed.
I think that you'd be having a very high bar in order to expect a brand new product to have as much content that could even come close to comparing to Wikipedia, which has been around for decades. It's like when Uber launched - of course it's coverage couldn't compare to taxi services around the world, it only started in san francisco, so you could only find a car in SF. Similarly here, Golden is focusing on only certain topics, e.g. clean meat and stuff listed on their homepage, for which they have a wider degree of coverage than Wikipedia. From those beachheads they can expand.
RE #4: "This is pretty slow considering how little you are doing" do you mean that the editor loads slowly, or the actual site topics?
@sidazhang Jude here from Golden. Thanks for caring and giving feedback.
1. I have not asked for reviews from my friends, but yes, some of them have given them. I am mostly interested to hear what product hunt core users think tbh. That is the value here to us to improve our product based on what you say.
2. We are opening the knowledge up on CC4.0 for the world. We are investing revenue from the tools we are making (not just with data) but the actual functions of the AI as well on private information and reinvesting this back into making more public data. The aim here is to build the largest open knowledge store and use company $$ to also accelerate this mission. Github has done well here, getting code more open sourced and charging companies for the functionality, reinvesting those $$ into a better product to get more things open sourced.
3. The content speed of production is increasing, new communities and products have to start somewhere. We have a lot of structured data and new articles are being built at ever increasing rates. We have focused in on various areas including biology, microbiome, blockchain and crypto, startups, AI technology, stem cell research and other very interesting areas. We won't yet everything but we will get there eventually. Have you tried the explore page and checking out the clusters we have been working on?
4. In terms of optimizations of speed, was the read or the getting into the editor slow for you? We have many moves to speed this up and make this the fastest experience here. Which country were you testing it from?
@iamedwardtay Jude from Golden here. We believe that search terms for these 10bn entities might make up 10-15% of search?
@iamedwardtay Google search results for most topics have a lot of noise and often just point you to Wikipedia or lower quality content on content aggregation platforms.
A platform like Golden makes it much easier to get quality content without having to manually search through pages of search results to find credible, comprehensive information. There are a bunch of different incentives at work in the eco-system of search and SEO (ex: where content aggregators trying to monetize their content game the Google ranking algorithm, often indirectly to the detriment of the consumer trying to find knowledge about a topic) that prevent Google from easily solving this problem themselves.
Not sure why would anyone come up with one more idea for one more way of organizing knowledge that isn't intuitive, or needed for that matter. Humans have already shown that all they need is a clean & simple search engine, not a busy cluster of things to navigate through. It actually reminds me the old AOL and Yahoo, which couldn't succeed for variety of reasons, and aren't used anymore.
Pros:Organized knowledge
Cons:Busy, gadgety, gimmicky, unclear, "unfree", unintuitive, unnecessary.
@konrad_lv That precise form of reasoning is behind lack of innovation. Why do we need what is already working?
@ma_cris_cordova I am all about innovation, as long as it's improving and optimizing something. You can't just create more complex GUI and call it "innovation". For example I never understood why someone invented Word Clouds and now they're distant past. And why Google or Craigslist haven't changed much in 10+ years? Because their usability is pretty optimal for their respective target audiences (CL is targeted towards people who prefer that 'oldish' type of UI). Though I think CL leaves room to be replaced with something nicer :-)
Super interesting product! Question- wikipedia is written by 90%+ men. What are you going to do to ensure diversity on your site?
@kendallhtucker Great question. This has been on my mind and others have brought this up, so we made specific moves to attempt to address this 1. We have excellent investors who are not male - Cyan from FF and Christina Brodbeck (first designer of youtube) as backers to advise us. Christina has backed me before and has a lot of insight into the UI/community side of this 2. We have Carla ex wikia in a high level position in our team 3. Updating/seeding the community early on will help address this 4. Proactively covering say female founders, scientists etc to get the content seed going. 5. Actively thinking about that issue WP has, why it happened, what could be done differently 6. Listening to the community on new ideas on how to solve the problem 7. Open to ideas from you here! We also added background diversity to the investor set as well, and have people helping from around the globe, as to map all knowledge, we have to do this globally with a wide set of people.
@kendallhtucker oh wow, where did you see this stat about the content being 90% written by men?
I can imagine it having some similarities to product hunt though, where I remember early on (I can't remember link to specific piece, but i think that it was techcrunch) a jouranlist said that PH team were strongly encouraging her to post products to PH as they wanted to increase the diversity of people submitting content, but she found it pushy as she DIDN'T WANT to be posting stuff to product hunt.
So then I'd say that maybe it makes sense to look at the reason why it's 90%+ men.
- If the reason is because Wikipedia is being run by an "old boy's club" with arbitrary rules and process and you need to 'know somebody' in order to be a good contributor there, then that might be a problem.
- but if it's equally accessible for anyone to submit content, but just some men might have more free time and be open to submitting content unpaid to Wikipedia, whilst not as many women might WANT to post to Wikipedia unpaid (it's kind of like a charitable service), then is that in fact an actual problem?
@_jacksmith https://en.wikipedia.org/wiki/Ge... Clearly it's a problem.. fewer female contributors means that Wikipedia has fewer articles about women or topics important to women. Also, "Wikipedia editors that publicly identify as women face harassment from other Wikipedia editors". 👍
@kendallhtucker ok. so with reference to an underlying reason, then Golden should be able to address that as a minimum
@_jacksmith @kendallhtucker This is a problem and we intend to fix it. It has been in our pitch from day 1. Here is a good example, https://golden.com/wiki/Clarice_...
I loved GOLDEN! https://golden.com/wiki/Juphy I'm a contributor right now ^^
Congratulations on absolutely fantastic launch - looking forward to where you can take it and whether you'll be able to overtake Wikipedia as main source of information one day. As long as you can keep the quality of the content up (and ideally improve it over time), I think you might definitely succeed. Good luck!
Looks amazing Jude. Looks like a smarter wikipedia(both for readers and editors). Plus a real business model so you don't have to bother people for money all the time. I can imagine Golden making a huge positive impact on society. Best of luck!
Looks awesome so far. Great idea. How do you aim to keep this apolitical? Certain things on Wikipedia become very political and give someone’s view of a person or topic based on their bias (be it liberal or conservative). How do you aim to try and remain neutral?
@iamjmw It is going to come down to the system we build/are building to bake in this neutrality into the design of the system throughout. Our business model is aligned with this objective and that is a powerful thing.
@judegomila The vision is brilliant. It have been wanting something like for so long.
Have you considered the long-term classroom applications? Textbooks in most countries are so outdated, and WP doesn't really have a good way contextualize any piece of content and link to Edx, research papers and more in a very intuitive way for an average user. Here - Golden can be the game-changer.
This is one of the more ambitious startups I've come across. Wikipedia is an amazing resource with deep defensibility.
What's been the biggest learning or surprise so far, @judegomila?
@rrhoover Jude from Golden here. Agreed that Wikipedia is one of the best things that has been created so far by humans - I use it all the time. The biggest learning surprise so far was the scale of lack of information compiled out there. When I started digging deeper the cardinality of entities we want to map came out to be approx 10bn which is around 1000x the article count on WP. I just kept running up against pages like this https://en.wikipedia.org/wiki/Cr... when it could be like this https://golden.com/wiki/Cryobact... or this https://en.wikipedia.org/wiki/Gi... when it could be this https://golden.com/wiki/Ginkgo_B... Too many topics were being deleted or not being covered fast enough or in enough depth. In terms of non product learning, building a company the second time around was just as hard as the first time round. The first time around the naivety is actually useful in that you take high risk, the second time around you have a couple of scars that make you want to feel more risk adverse. Building products that people use daily is a very very hard thing to do. Another major learning part is, your team is everything, so hire the best (which doesn't mean most experienced necessarily but more so motivated by the mission and faster learners).
@rrhoover @judegomila But also this https://golden.com/wiki/China when it could be this https://en.wikipedia.org/wiki/China
How do you plan to solve the problem of scale? Wikipedia has millions of contributors motivated by it's non-profit nature. How do you expect to get 1000x more contributions in while keeping profits for yourself?
@rrhoover @judegomila @freediver "Wikipedia has millions of contributors motivated by it's non-profit nature" - that said, some moderators do take advantage of it with a behind the scenes black market economy (which I'd be interested to learn the size of), see: https://t.co/l6VJbgDT6H
@rrhoover @freediver Jude from Golden here. I think the motivation is more around openness of the topic pages and Golden is running on CC4.0 which is the important part of the core motivation - to get everyone access to the pages and knowledge. I don't expect to get 1000x more contributions, we are writing software for leverage so that 1 unit of time spent on Golden can produce 100-1000x more. There are some tasks as well that can be fully automated, over time we see this automation increasing. Additionally, we actually have been reinvesting the $$ into building a better product and collecting more information to open source. Github has done a great job in allowing company $$ to be reinvested back into the product so that more code can be open sourced ultimately. To solve the problem of scale we wrote up a blog post above to describe how we will automate and semi automate much of the process.
Summer Bod 2020
Really enjoyed the tools they provide to collect and source the knowledge.
Pros:So much easier to use than Wikipedia
Cons:Still small # of articles.