swframe

Giving up usually doesn't help. Just taking time to identify the root causes often leads to a simple and better solution.

Unfortunately, I am rarely able to convince others to follow along.

My company is extremely successful and our spectacular failures (often discussed on HN) pales in comparison to our success. I worked for several groups that lost multi-billion dollar markets because they were unwilling to make simple and obvious changes. For example, don't focus on our success and claim victory; look at our failures and make corrective actions. Seems pretty simple but maybe due to the innovator's dilemma, no one seems to care.

The CuriousAi TAG and LADDER research looks interesting. They claim same accuracy with much fewer labeled data.

I've built similar platforms. Naturally, I think they are awesome. They are a core part of very valuable products like SAP, Salesforce, etc. You will get a lot of push back from high developers who prefer direct access to the native api or html. Please don't let them discourage you. My response to them is that they can build reusable components that are driven by your json layer (win-win). There are enough developers who will benefit. Try to get funding sooner than later. I waited too long. You can't keep up building this by yourself.

We focus on the mating but the real challenge is keeping relationships interesting when we're not mating.

What if large companies built on-campus employee housing on top of their parking lots and moved the parking lots below ground? They could make a "corporate" city by adding a mall, hospital, child care/schooling. Imagine being able to have lunch with your family everyday; having a 15 minute walk to work.

Is there a link to get the code?

I used to give Dune to my teenage relatives and it usually turned them into avid readers; they would thank me later. I noticed that it doesn't hold the interests of the recent teenagers in the family; it is hard to get them to start and even harder to get them to finish. I think competition from the internet, youtube and cellphone is just too strong. That is horrible because developing a passion for reading helps lot in highschool/college/life. I've offered to pay them $100 and they still won't read it!

Agreed, I made gross assumptions based on my crazy background and experiences. I grew up in public housing (80% welfare families) in East New York (Brooklyn) during the 80s when that tiny (crack infested) neighborhood had the highest murder rate in all of NYC. For the last 3 years, I currently fund the education, teach and mentor several people in the 3rd world who make less than $10/day. The poor people I grew with and I now try to help are ...

1. Smarter than me but don't know how to end the cycle of poverty (one had a baby recently at age of 19 and both parents are unemployed).

2. Are not lazy but work in dead-end 60-hour/week jobs where they are not learning the skills that allow them to make enough to escape poverty and have no time to learn new skills.

3. Are not criminals but have accepted their fate and are not working to change their neighborhoods to reduce the temptation of their kids to enter a life of crime.

4. Are not drug users but drug related crime and extreme violence surrounds them (they have seen dead bodies in the streets on their way to work).

Other kinds of poverty exists. My proposal should be adapted to work in the situations I've not experienced.

"Applied" education is the best path for the poor. Don't teach the 3-Rs. Teach them how to fix their problems: 1) fight the cycle of poverty 2) weaken gangs 3) reduce crime and drug usage 4) renovate local housing, schools, and infrastructure 5) create and run local businesses 6) employee all teenagers after school and during the summers, etc.

Once they have pass those classes, they can learn the other stuff.

And this is how we should pay for it. Sell treasury bonds like we would to raise funds to improve our national infrastructure. In other words, invest in poor people in such a way that you can repay the investment from their future taxes. Specifically, pay poor people to work their ass off to fight poverty. (No person should be paid to do nothing; that is a horrible idea.) Allow the wealthy who want to reduce their taxes to keep $x in taxes for every $y they invest in hiring poor people to fight the causes of poverty. In fact, we should create poverty fighting companies who complete and measure their success by the quality of life improvements they cause. The higher performing companies lower the taxes of their investors.

(This isn't even expensive. Paying a teenager to avoid a life of crime costs $1000s/year and putting them in jail costs $10,000s/year. Furthermore, providing teenagers with IUDs costs $100s and saves $1000s.)

The US has a horrible human rights history. The country started with genocide of native americans. It enslaved and murdered millions of africans. It still has very racist tendencies. It killed several million koreans, vietnamese and cambodians (indirectly). Then there is Iraq and Afghanistan. There are a lot of really horrible things the US did that is on par with the Nazis and the worse we've seen from many other countries. The US is good at hiding or ignoring its atrocities. I don't think China is better or worse. But I don't see why it should be vilified when the US (and many other 1st world countries) have done much worse things to each other and to the 3rd world. In fact, I think all 3rd world countries should grow their internal markets in a similar way until they can become 1st world countries.

Do you know of a human pose estimation from 2d images service/library? I've seen papers about it. I would like to try it out.

Can it see proteins, cell walls or viruses?

What an odd coincidence? I had a similar idea yesterday. Except, I think your solution should be more general; A site that lets anyone pay to talk to an expert on any topic. The user should post their question and the experts compete to be selected; it is important that only the experts that care about the question are selected. Kind of like uber for conversations. I don't think it should be a TaskRabbit clone. The experts only provide information; they don't do the work.

For example, I want to talk to someone about selecting an alibaba manufacturer for a specific niche. I also want to talk to someone about how to get products into retails stores. And so on.

There are many related services. For example, you can have translators connect people. I think there is also a need for translators for people with poorly formed ideas; the translators would communicate with the experts who don't want to deal with those people directly.

Previously, Martin Shkreli shorted drug companies and tried to convince regulators to blocked their pending drugs. I wonder if he is shorting the drug stocks again but this time he is using a new technique to destabilize the market. So the steps would be:

1. Short stocks in a vital niche market

2. Buy an inexpensive producer of an important product, jack up prices to cause outrage and a strong government response

3. Wait for the market to sell off stocks in the niche market

4. Collect the profits from 1.

5. Dump the producer purchased in step 2.