935 karma · joined May 7, 2014
Another Vitamin whose deficiency cause irreversible damage is B12. Folks, get both of these checked.
This confirm what I have suspected since almost a decade. I have been taking 5k Vitamin D per day since one of my cousin had serious health issues because of low vitamin D for 3 years and as soon as she started having 5k IU per day, they disappeared within a month.
Saudi Arabia is trying to buy all these assets to secure their future. Returns will likely be reinvested or stashed abroad!
However with the dismal future of oil based economies and the unstable situation in Middle East - only scenario their would be able to retain these assets would be if they are somehow able to transition into a functioning (hopefully democratic), educated and productive society - which would be awesome! It's better for them and the world. I would love it if they are able to use this money to achieve that!
If not they would descend into chaos and all assets would likely be seized by whichever country hosts them. E.g. Gaddafi!
[1] https://www.theguardian.com/world/2011/feb/28/us-treasury-bl...
[2] https://www.theguardian.com/world/2012/mar/28/gaddafi-assets...
[3] https://www.theguardian.com/world/2011/feb/27/gaddafi-family...
There is no future where Uber is profitable and there isn't someone with deep pockets who wants to undercut them starting with their most profitable markets.
Most riders don't have much customer loyalty, they just want the cheapest price. Drivers don't mind driving for multiple services.
When a deep pocketed competitor launches with zero market share, they can compete with lower prices. If Uber responds too early, Uber would need to cut down prices on their full market share. If Uber waits too long, the competitor gets more traction.
In the end it is a zero sum game for Uber, unless they could have used their money and focus to create Autonomous Driving Tech. Not sure if that is going to happen now! Their new CEO's mandate is to do the IPO and get investors their return ASAP.
My prediction:
In the next 2 years, Amazon will enter as an Uber competitor starting with Seattle and then slowly expand to Uber's most profitable markets.
Any bets :)
The original comment was wondering why should Uber invest in building Autonomous Driving at all.
And I am suggesting that once we have true autonomous driving (all conditions), and if Uber doesn't have it - Uber has not reason to exist at all.
If the tech is indeed safer vs human, it would also have an added benefit of not having safety risks similar to Uber where some Uber drivers have assaulted their riders.
Benefits of Autonomous drones:
1. Do not need to deal other non autonomous devices while in flight. Cars needs to deal with human, other cars driven by humans.
2. Do not need to follow existing road traffic rules.
3. Lot of paths available. Cars are limited to existing roads. Drones can have a lot of flight paths.
I understand that there are other challenges - more energy required, less safe in case of an accident, flight regulations.
But still. It seems like if the technology is there, it would be an easier transition with Drones vs Cars.
Right now an Uber driver who is probably online 60 hours a week, needs to support a car priced around 25k (amortized over 5 years), gas (which is more expensive than electricity), maintenance and 50k-70k/yr driver income. Assuming lower end we are looking at 5k car + 10k fuel + 5k maintenance + 50k driver income. So cost to support a driving car is 70k/yr for lower hours of operation.
So we are looking at a 5x cost and lesser hours of operation with human driving vs autonomous driving.
These are off the napkin numbers so please don't take them literally. It is meant to highlight that driver income is the biggest factor.
Whoever has self driving tech, they can start undercutting Uber with the most profitable markets (NYC, SF, Boston, LA etc) and expand from there.
I feel that we are at least 5-8 years away from main stream self driving. But I could be wrong.
Two party networks are much more difficult to solve than one party network problems. Because with two party networks you have a chicken and egg problem.
Anyone who has solved self driving, has near unlimited scale on the driver side. Their cars can autonomously be present to respond to peak demand - e.g having thousands of cars at Burning Man.
In the autonomous car world, Uber has no place if it doesn't have that tech. There is almost no reason for the company which has self driving tech to license it to Uber, when they are easily run it themselves. The only value Uber has right now is because it has crossed the critical mass on the two party network. And some people thought they could solve self driving cars.
OTOH, a global two party network (e.g Airbnb) is exponentially tougher than a city level two party network. Which is why Airbnb has almost no competition (except some local players in China), while Uber has competition everywhere. In fact, Uber has already lost and conceded a few major markets (China, Russia).
Can Uber - even sustain what they already have. I have my doubts and theories, but that's for another day.
I was also worried about random code getting downloaded to my machine and getting executed.
However from Uber's point of view - cost of 24 hours to rental company vs cost of 24 hours of taxi+driver would probably be a good approximation to get an idea of the order of saving which might make owing self driving cars feasible for Uber vs their current business model.
On the other hand, I feel that we will not have self driving cars for a really long time. Instead I believe we will have flying cars before that, who will solve the last mile problem. They will take you from the house to the hub and hub to the final destination. Buses can transport people between hubs.
With self driving cars you need to deal with existing regulations, traffic rules, other drives on road, people jaywalking, animals running into you.
However with flying cars - you literally have exponentially more roads in air (which don't even have to be built). No existing regulations (unless at risk of being in flight path). Technology is already there (make drones bigger). These drones will communicate using MESH networks leading to lesser risk of collisions.
I also feel that Elon Musk is doing something really smart with The Boring Company. If he can dramatically reduce the cost of building tunnels (like he did with space travel) - he could build a lot of underground roads at a lower cost and faster pace than roads on surface. No need to deal with buying out land and getting approvals.
Imagine if these tunnels are exclusive to Tesla cars, who can auto-drive these tunnels for long distances. No need to worry about other cars, animals, rain, snow, traffic regulations. All the problems that auto driving cars are facing now, will go away. Multiple Tesla cars will communicate with each other using MESH network. Teslas could be driving in these tunnels at 150+ mph and discharged cars could be swapped with charged cars on the way. Drivers might need to take over minimally just for the last mile.
If a tunnel like this could be built between SF to LA. It could be faster door to door vs flight.
To get started, a tunnel from SF to really South Bay could have an huge impact on the commute.
I do not have much experience of these industries. So these are my hypotheses. Would love to know what others think!
Given a choice, Uber would love to be the first to have self driving cars. Once they do that their business model scales predictably with capital.
Whoever gets self driving technology right, will own all transportation and logistics - which is huge!
However, I doubt they can do it. The quality of engineering talent is subpar compared to Google. Given that they couldn't compete with R&D they probably resorted to other means.
http://www.cic.gc.ca/english/express-entry/
It's called Express Entry which gives Permanent Residency. If you are selected, you get PR before you land in Canada. The reason I said Green Card is that this is the term most people on this forum are most comfortable with. What I had meant is PR.
Did you know that any employee at these companies can claim to be a Multi national manager and get a Green Card within six months. That person might not have the qualification to pass high school.
They knowingly bring people with fake degrees. They knowingly break the laws. They claim that people are working at their site while they are really working at the client site. The list goes on and on.
The consulting company used to send really low quality candidate. The kind who might not clear high school maths and have some worthless degree from a remote college in India. Then I asked them to get their best people. Their best people were pretty bad. I mean the kind who would get rejected on telephonic screen within 10 minutes. Eventually the model that worked is that I would find great candidates and then this company would bring them on contract (some people prefer working on contract vs full time).
The amount of money waste on such consultancy projects was unbelievable. Companies prefer this because of some tax savings of CapEx vs OpEx. There would be a huge team of people doing near clerical work. I am confident that there are other citizens available who can do that job, at probably an higher cost.
Only of my friends, who has a profile very similar to mine just moved to Canada. He was given a Green Card before ever setting his foot there.
Instead US is getting people with really low quality education or fake degree, who are brought here by staffing company. They get paid really low salaries and displace American workers. The current system is encouraging and supporting these companies in taking advantage of the H1B system.
Because there is a lottery every year, best people from across the world do not get a chance to come here on H1 because the quota gets full. If someone decides to move to US on Apr 1st in 2017, they need to wait till Apr 2018 to see if they win the lottery. Even if they win, they can't join till Oct 1st, 2018.
Back in the days, when the quota didn't get full, people could apply at any time and join soon after. If you applied between Oct to Apr, you could join right away. If you applied between Apr-Oct, you could join in October.
It is in US's interest more than any one else to fix the current H1 system. Best among whatever Indians are stuck here in limbo will eventually move somewhere else. Best from whole world do not want to move here on H1 because of the uncertainty. This is not what the system was designed more.
There is a lot more to the story why big companies like contracting to H1 and L1 workers (tax benefits of Capex vs Opex). That story is for another day.
It took a lot of hard work and luck to get me to overcome this. So I would say that my life improved in-spite of H1. I was able to found a startup (and successfully exit) while on H1, which is extremely rare.
Around 2010, the Green Card wait for Indians went from 3-5 years to 20+ years. Many people who moved before or around that time got invested here (bought a house, grew family, kids going to school) and in a way are stuck in a limbo. Many of them are smart and hard working people, who used to have dreams of building companies. But after a decade in limbo, all they talk about is the next H1 renewal and the uncertainties around that.
Now, the best from India are instead choosing to stay in India and build startups there. While the current system is allowing staffing companies bring subpar talent which displaces good American jobs. So yes the H1B system is broken and it needs fixing.
If I had stayed, I would have had different set of opportunities in India. Difficult to predict whether they would have been better or worse!
And it is easy to say "that's fine" - when you are not living the life with constraints.
Maybe your friend wanted to do a startup where he owned the majority voting control in the startup. You can't do it on H1. Maybe he wanted to go meet his family but couldn't because his visa renewal was going on. Maybe he wanted a great opportunity, but the company wasn't setup to do H1 paperwork.
H1 was expected to be a temporary state. If you are on H1 and not from India or China you can get a Green Card within 1 year. Even Chinese nationals can get it in 4-5 years.
I switched 5 times in 6 years. Joined LinkedIn, left to build a startup (where I couldn't have majority stake), sold it to Dropbox, built and incubator at Verizon, doing another startup. All this while I was doing fine financially. However I felt really suffocated by the lack of freedom.
Finally I am not on H1 and now I can own a controlling stake in my startup. If I wasn't on H1 earlier, I would have made different career choices all this while.
H1 system is broken. Companies all over the spectrum is using it to exploit employees. It's time to fix it.
And if the intention is to solve the problem of jobs moving to India, create a law designed to fix that. H1 wasn't designed to fix this problem. Let it do what it was designed to do!
A small part of the visas (less than 10%) is used in the long tail by companies like Google and Facebook. They jobs pay relatively well. However because of difficulty in job mobility, the immigrant employees lose on negotiation and aren't paid we well as they could have gotten paid. At least these jobs do not displace American workers. The above companies have a lot of unfilled openings and if an American worker is available with the right skills, they would hire that person.
However current reality is that very few startups employ H1. I founded a startup when I was on H1 hired someone else on H1 and it was a pain to go through all paperwork. Right now, if the potential employee doesn't already have an H1 - the startup will need to apply for H1 lottery where odds are 1/4 and if they win it wait for 6 months before the employees can join.
Recently I built an Incubator where the intention was to pay market salaries while people try out ideas. And once the idea gets traction - raise funding and spin it out. Everyone on H1 refused to join the spin out because they didn't want to deal with the pains of being on H1 while dealing with uncertainties of a startup.
I hope that these regulations stop the abuse happening on H1. This will help in two ways: 1. No lottery - so people who want to move to US have a predictable path. 2. Better path for Indians to get Green Card. Right now the delay is insane (20+ years). Once they get Green Cards they can participate with more freedom.
I have suffered because of this. I spent 6 years on H1 with no end in sight. It would take at least 20 years for Indians to get a Green Card. Till then, you have to do do insane amount of paperwork, deal with lot of uncertainty when we you are up for renewal and are in a limbo state.
Most of my friends who are really smart and are in US, have resigned to their fate and given up on their entrepreneurial aspirations.
If people think that companies like Facebook and Google are not exploiting people, they are mistaken. Even though the wages that they pay are much higher - H1 creates lack of job mobility, within that class of employees. I worked with someone really smart in an earlier startup in India. He spent 7 years in Microsoft because it seemed like he will get Green Card next year.
I worked with someone senior at FWD.us and as per that person the intention of FWD.us is to get more H1 and not Green Cards.
I have lived through this and hopefully out of this cycle. However the best from India aren't moving to US anymore. They now know about the uncertainties in their life because of H1 and are choosing to stay behind in India to build startups there.
Big consultancy companies (TCS, Infosys, Cognizant etc) are abusing the system. I hope someone puts and end to this.
Edited: formatting
If someone is joining a startup at an early stage, work like it is your own baby. You are spending your most valuable currency by being here - which is your time!
With time, hopefully, you will get rewarded for your effort and results. You will likely not get a founder level say and stake but your role should gradually expand. In the end founders take almost the same risk as the first employee, but get a lot more equity because founder was there at the start. Join a startup as an early employee only if you are able to accept this fact - otherwise you are setting yourself up for a lot of resentment and pain.
This above advise applied to me too for the first two years of my employment, before I started working on the ultimate successful product. We were building an offline search engine and as a part of that I build distributed file systems, crawled a billion pages and I was appropriately rewarded with career growth.
However with the new social product that I built, the situation was dramatically different. For that product I was there from day 0. I was owning the whole engineering (sole engineer at start and head engineering till I left) and significant portion of product. Till we got 5M users, I had a co-founder level say (but not the stake of course) and I worked with a co-founder level of involvement and dedication till that time. However once we got 5M users, multiple levels above me started doing meetings and taking critical decisions without involving me. I still had a co-founder level passion and it hurt to see the product flounder and being unable to do anything about it.
So I was in a uniquely bad situation and it is not common to be in this situation. However if someone really is and there is no way to get a founder level control - move out as soon as you can. And this is what I did - albeit 2 years too late.
As I look back, if we had used RoR or Python we might have moved faster with no negative impact on scale. We would have spent more money on extra servers though!
If someone cannot scale your product which has adoption, they likely don't have insights, problem definition, capabilities, skills and resources to do that when they launch.
If I could edit my original post, I would put a rider related to capable team for service not dying.
My post says 50M+ users and 1B+ message on a day. Almost 50% daily active users. And this was early 2000 hardware. Last year WhatsApp was doing 60 billion messages per day (60x of what we did).
However our messages had a lot more logic, because we delivered on unreliable Telco texting pipes. Pipes had a set throughput. Some pipes delivered to just some geographies. They would randomly fail downstream so you had to do health monitoring and management. Some pipes cost per message and some cost based on throughput. So pick in real time for the lowest cost. Different messages had different priority. In summary - there was a lot more business logic per message sent.
I was of course very inexperienced when I started. I was just two years out of college. However in the end the product was 50+ services running on 200+ servers. This experience helped me scale user communication infra at Dropbox to 100x scale within 3 months.
Your message was flame-bait at best and didn't seem to contribute to the discussion. You were passing judgements based on incomplete information. You could have asked nicely and still gotten a response.
Edit: fixed typo
TLDR: Depends a lot on individual situation. Pick the language you are fastest and most comfortable with.
Background:
SMS Gupshup - 6 years - Java and some C++ - Built distributed filesystems. Crawled 1B pages with early 2000 hardware. Built map-reduce framework before Hadoop came. Built a search engine on top of it. Built an infra which would send 1B+ text messages with prioritization and monitoring. Social network with 50M+ users.
LinkedIn main stack - ~ 1 year - Java
LinkedIn mobile stack - ~ 1 year - Javascript - Was one of the early engineers.
Founder Startup - 1.5 years - Clojure
Dropbox - ~ 1 year - Python
Head Technology Tech incubator - 1.5 years - Node.js
As you can see my experience has been all over the spectrum. While primarily using one language, I would keep experimenting with other languages like Scala, Go.
Following are my learnings and the reasonings:
I love Clojure. Love it so much that I feel like quitting everything go to a mountain cabin and just code Clojure. However I would most likely not use Clojure for my startup. The learning curve is really-really steep. It would make it very difficult to me to build a team. It's really fun if you understand it. But first few weeks for a new person might be a nightmare and very emasculating. So even though I really love it, my current strategy is to bring Clojure patterns to other languages.
User facing logic
I would not use a statically typed Object Oriented language for any user facing logic. User facing logic and behavior tends to be very fluid and I feel that it gets strangled by the constraints of statically typed languages. One you start building things with Class, Inheritance and Polymorphism etc - soon these patterns start driving business logic rather than the other way around. Because of this reason you will see most arcane and slow moving popular sites today are Java based - (LinkedIn, Yahoo, Amazon, Ebay) vs (Facebook, Twitter, Instagram, Pinterest). For this purpose I like using Dynamically typed languages building code in as functional way as possible. Pick Node.js, Python, Ruby - whatever you are comfortable with.
In some cases people recommend Statically types languages for catching errors because of type safety etc. However I would solve that problem completely with test suite rather than solving it part of the way using language features while being forced to pick a statically types language.
Another big benefit of dynamically types languages is ability to just create a dictionary and start using it. Most CRUD apps work with JSON request and response and I would rather just dynamically pass objects, rather than build a class for the request and response objects for each and every route. This was a nightmare to create and maintain in Scala and Java.
Fast developing Infrastructure
E.g. Hadoop, Hive, Distributed FS or Queue. 5 years back I would have built the in Java. However I would use Go for this today. However you couldn't go wrong either way. Pick what you are most comfortable with.
There is decent amount of talent available. These languages are easy to learn. And you get close to C level performance if you built it right. Development is fast paced.
Mission critical performance - File System or Database
Most likely C, because you want to squeeze out last of the performance. However be ready for really slow development cycles. And most smart kids out of college wouldn't know C. So you will need to set aside some time till they are really productive.