ericxtang

Hey - Eric from Daydream here. Thanks for linking our project! Check out this tutorial video that Vibor made - https://www.youtube.com/watch?v=cW3ZGxZNX0g

We've been working on an alternative infrastructure and saves up to 80% on transcoding & delivery costmore affordable solution at Livepeer Studio (https://livepeer.studio/).

It uses un-utilized infrastructure around the world and incentivizes independent network operators to join the network (kind of like a 2-sided marketplace for video-specific compute).

Please sign up for a free account and check it out! We'd love to get your feedback

The underlying technology of Web3 (distributed consensus building) is quite different from that of p2p (networking and information sharing).

While some of the criticisms of web3 are true, most of them are situations caused by the early phase we are in. Moxie's post surfaces the need for better decentralized infrastructure. The hyper financialization (everything is monetized) is caused by lack of identity solution and limited blockspace in L1 chains. These things are being built now, and will come online in the next few years. Transaction fees will become cheaper, applications will be more decentralized.

The core innovation is digital scarcity and crypto-enabled data ownership. For the first time in history, we can represent value without a middle-man. This completely changes the power dynamic of platforms vs. users, and makes value creation digital-native. Web3 has an opportunity to address some of the biggest problems with today's internet services like surveillance capitalism and censorship.

Yeah the bandwidth tax on the cloud makes some use cases impossible.

What do you do instead? We use Ansible to bootstrap k3s. It works, but we have to build a lot more stuff related to monitoring, alerting, routing, etc.

Livepeer is hiring for a full stack video engineer.

Livepeer is a decentralized video streaming network, and provides a live streaming service that's more reliable and scalable, with over 10x cost savings.

Check out the product at livepeer.com by signing up for a free account.

If you are interested, please apply at: https://livepeer.com/jobs/full-stack-video-engineer

Hi wmf - I work on Livepeer, happy to clarify on some of the skepticism here.

Livepeer uses cryptographic signatures to prove authenticity of the video transcoding results, and that signature can be used as proof to punish cheaters (if you signed the result, you are responsible for the correctness - and since it's easy to validate whether the transcoding is done correctly, anyone can use that signature to punish you by taking away your stake). If no one cheats, no one gets punished. Since the cheater doesn't know how many people are monitoring the results, their best strategy is to be honest.

Very true. Engineers like that are extremely rare, and usually have serious golden handcuffs. However, there are a lot more people with similar mindset but limited programming experience. With the right culture, those are the usually the people who can make a big different at a startup.

I've heard Google has a special process, unfortunately even if they make it public, it would only help startups in a very limited scope. The hiring requires are different, especially when it comes to a lot of the "soft skills".

top 5% is a very relative description. There is no hard measure, and everyone have a different standard on what they are "measuring in their head". Our perceptions are extremely biased by our own experiences. 90% of the time I really have to get to know the candidate before making any decisions. Of course the biggest constraint is available qualified candidates at the time. That's why hiring is so hard, and people who are really good at it tend to be veterans who have been in the industry for a long time.

What about things like Titanium or RubyMotion? They compile to native and allow you to have a "unified" code base.

I've captured a lot of my thoughts about this here: http://www.erictang.org/process/2012/02/16/brainstorming/

Square does an amazing job at solving both a hardware and a software problem in just 2 and half years. Pretty impressive.

Classroom setting is really not the best setting for studying math. Studying new materials in math needs complete concentration, a 2-3 minutes mind-wondering can cost your the entire period, since new concepts/abstractions in math build on top of each other.

It's very difficult to concentrate for a 40 min period, especially considering students just came from other classes that could be just as mentally challenging.

I think Khan academy does a good job at this. A 15-minute long video is a much more consumable chunk of knowledge.

It's super difficult to find a startup PM at the beginning, partially because it's super difficult to find another person who has the same product vision. It's much easier for the founder with strong conviction to see the product through, at least to a stage where it's been proven by the market.

Like any kind of spamming, I'm sure it will be stopped, especially if it's not even automated.

Do you have a particular algorithm when selecting pictures?

Khan academy is a huge step forward, delivering key concepts in video forms. However, this approach only goes so far. Human interaction becomes important when higher-level concepts are introduced. For example, I can't imagine getting through my O.S course without the help of TAs, office hours, etc. Recreating that experience on-demand would be the holy grail.

We've built an API using ruby on rails. I highly recommend the book written by Paul Dix if that's what you are using.

http://www.amazon.com/Service-Oriented-Design-Rails-Addison-...

I think people you know are the most important resource in terms of finding a business-savvy co-founder. A lot of VCs/angel investors are willing to help smart hackers find co-founders or even interesting projects. Also, going to marketing/product related meetups is a good way to network and meet new potential co-founders.

I think it different in each city. I live in New York, and this city is in dying need for more engineering talents to feed its fast-growing startup scene.

Being a hacker, the last thing you wanna hear is some finance dude pitching you an idea they had when they were crunching numbers on an excel spreadsheet at their 2nd year analyst job.

You have be product-savvy, you have to have respect for the hacker community/culture, and you have to be super persistent.

CUDA has come a long way since it's birth as a student project.

You can totally just set up a mirror in the U.S to access any websites. There are also services out there to get around the firewall.

Really fantastic piece of software. It's about time we move away from using java wrappers for NLP stuff. Anyone know a similar project in Ruby?

There are plenty of successful enterprise software providers and there always will be. It really depends on the expertise of the founding team. The product just has to make sense. Some companies stop at the point when the purchase decision make likes the product and stops innovating (or they call move towards scaling, i.e. devote all resources to sales), and that's when they lose their edge.

Anyone know what the "bottlenecks" are in YC? I assume it has to do with the number of advisors? Also, I wonder if the growth in number means decrease in quality (or experience) for the entrepreneurs.