Ask HN: If you had 100 talented people for 100 days, what would you work on?

[yaronhadad]

Imagine Jeff Bezos offered you 100 of his top people (engineers, designers, product ...) for 100 days and told you that you could work on whatever you wanted. The only condition is that at the end of the 100 days you need to start paying salaries in order to retain these people.

What would you work on?

[rustler]

A portable personal Internet server that runs as a turnkey no-user-maintenance image on AWS, Digital Ocean, and OpenStack. It hosts your email, federated social networking and chat, file sharing and ad-blocking web proxy. The user experience is pure Internet with no ads, surveillance or vendor lock-in, all for $6/mo (hosting + DNS).

[l0b0]

This is an idea I've been toying with several times, but it never gets beyond the thinking stage because of the obstacles interacting with the rest of the world: getting at least two cloud provider accounts (one for regular operations and one for secure off-site backups, ideally with a different provider), estimating costs (because good luck actually calculating them), buying a domain and transferring it to the cloud provider, the intricacies of configuring email to not be black-holed immediately by other providers (something I only know from rumours), and I'm sure a host of others. Even an effort to make some of these things slightly easier would be most welcome, but I don't think 10,000 work days is enough to get something like this to a "turnkey" level. It would be absolutely the best to be proven wrong, though!

[rustler]

Maybe we could get it "mostly" turnkey just for email, and charge $10/mo to fund 2-3 full-time devops to maintain and extend it.

[ttul]

Oh yes please. Seconded.

[anoncoward111]

Instead of worrying about the practicality of this, I'll answer your question directly.

I would aim to create as many blueprints as possible for tools that can help more people live and work remotely and sustainably.

There is no greater threat to human health, human happiness, and human productivity, than there is in meaningless desk work and commuting.

[stevekemp]

You think the ability to work remotely is a more significant health-benefit than tackling water-shortages, obesity, drugs, and climate change?

It is OK to have a niche, and to use hyperbole, but when it comes to human health and happiness I'm certain there are more significant issues which are real, and growing, than where somebody works (if they can, must, or even do have a job).

[anoncoward111]

The problems you have listed would immediately cease if US and Chinese corporations had any ethics whatsoever.

Obesity is caused by stress eating, cheap junk food, marketing, and a sedentary lifestyle due to a desk job. 2 billion people in South/East/SouthEast Asia live long lives because they are outside the influence of the Western diet.

When you say drugs, I dont know if you mean medicine or heroin. Most medicine is cheap. Pharmaceutical companies and insurance companies hold a cartel over its pricing. As for illegal drugs, decriminalize it like portugal and your problems will cease.

Climate change should be obvious. If everyone stopped consuming the wasteful products US corps market as disposable (gas, cars, meat, etc), we would cut our greenhouse emissions overnight.

All of these problems would end if corporations were somehow forced to stop. I doubt anyone will ever acquire the political momentum to do that.

[ghosterrific]

Yes.

The problems you describe above would be greatly reduced.

- Less fuel consumption means less water use, less pollution

- Less wasted time commuting means more time to relax and less stress eating, more time to prepare better food and better sleep

The ability and effectiveness of working from home will have compound improvements to the other specific problems you listed.

Imagine if every worker had a couple extra hours to sleep better, eat better and time with family to solve the problems above.

[rustler]

I would focus on patterns for remote collaboration (structured and searchable chat, VR hangouts) combined with old-fashioned face time (monthly or quarterly in-person activities).

[songzme]

I would have them train 2 people (each) in their field to be as good (if not better) than they are. At the end of 100 days, I'll have 200 people trained by Jeff Bezo's top people. These 200 people will be tasked with building a learning system / school that teaches these skills to underprivileged communities for free. Each school would be named after each of the original 100 top people.

I'll keep tabs on each of the 100 top people and ask them for donations every year to keep their schools alive. If they don't (and we ran out of money), we'll close down the schools.

[warent]

Ask all 100 of them the same question, but they get 4 people instead of 100, collect the top 20 best results, give those 20 folks leadership over a team of 4 people, making a total of 20 teams of 5.

Statistically at least one or two of those teams is likely to come up with a highly profitable idea.

[pvinis]

I see only one flaw. Who choses these top 20 results? Maybe he cannot imaging how something could be profitable. I'm kidding but it's an interesting thought. Thanks for that.

[westoncb]

First off, I'm crossing my fingers it was actually a drunk Jeff Bezos who posted this :D

I hit on a new general scheme for structuring programming tools and languages that doesn't use parsing, and I'd love to try building a language and editor on it. I've seen other people working on these syntax-free, no parsing schemes, but I've yet to see one whose fundamental structure is simple enough to be captured by a simple diagram, or just a few sentences; and that's what I'm lookin' at here, which leaves me endlessly curious about the potential significance of this thing (including the potential for its significance to be nothing!)—but I haven't had a chance yet to give it a real test.

So if I had the resources I'd do a bunch of iterations on it—likely with ~10 teams working on parallel iterations if I had 100 people. Actually, I'd start with just one team of 10 working on it, and if it continued looking promising after the first iteration, and add more teams in proportion to the perceived promise. Meanwhile, I've got plenty other (over)ambitious projects I'd like to have help with :)

[ericjang]

Auction the team as a resource to the highest bidder. I'm not positioned to deploy this resource effectively, so I'm likely to benefit most by trading it to a buyer who can't convert money into talent fast enough.

The team is much more valuable to someone who can afford to pay long-term salaries and needs this talent resource quickly on an up-front basis (Tesla, Uber, a startup trying to crush their competitor, etc.)

[gitgud]

It seems selfish on the surface, but to me this is one of the best answers to the question.

Why waste valuable resources, when someone more capable could utilize them.

[ericjang]

Let someone else waste the valuable resource :)

[chipuni]

The people, by themselves, wouldn't be enough. Let's assume we have the people AND the kind of resources 100 people working at Amazon would have.

Then I would have them work on a multi-pronged project to get a higher percent of people in the US voting.

One team would be campaigning in all states against laws that make it hard to vote.

Another team would be finding the most cost-effective ways to register voters, educate voters accurately about what each candidate represents and has done, and get them to the polls.

[r00fus]

Of course there are going to be folks who claim that it's too political. There is one party that seems to benefit from preventing voting participation and fairness...

[raverbashing]

I would have them prototype a new computer language + ide for it

This language would have usability in mind as well as throwing out established assumptions. For example, that source code should be ascii text. Or that anyone should decide between tabs/spaces.

[adamnemecek]

Not enough people. Languages are hard as well as bad for making money. 100 days is too few.

[milkytron]

There's the possibility that money isn't his top priority.

[raverbashing]

You're right about the making money part

[frozenport]

Assuming that syntax is the hardest part of programming?

[raverbashing]

No, it would be only one of the aspects.

[peterwwillis]

Oh, that's easy.

Develop a non profit program to bring tech skills to disadvantaged communities so they can get basic remote work jobs and put food on the table. Key word: 'bring'.

It wouldn't make any money at all, but it might change some lives.

[DanielBMarkham]

I'd let them go. Keep the best 5. Or have a contest where you narrow the field down to 5.

100 people fresh for 100 days is a disaster. If you're lucky they'd all ignore one another, group into small teams, and find their own way to do cool stuff.

3-people teams can grow into 100-people teams that do some awesome things. 100-people teams dropped in from the sky are trouble. I've seen it done. It's rarely productive.

[singularity2001]

It's all a question of management. As someone said: build 20 teams of 5.

[DanielBMarkham]

Yes, it might work great as 20 teams of 5. No, not probably the way most people would do it.

It's not a matter of management, it's a matter of self-organization. People are not fungible like money. They are also not robots. So you just can't "manage" them or slice them into various groups and expect it to work well. You also can't take one giant problem and create 20 teams and expect them all to decompose it and solve it optimally without a lot of guidance. The things that work are counter-intuitive and the things that seem like they should work actually create a lot of friction and waste.

Put differently, there are a ton of startups that never made the leap from a 5-person team to a 100-person small dev shop. There are really good reasons for that -- and those are people who have been working together for years.

[patrickmay]

Create a full replacement for Javascript based on Scheme and integrate it into Firefox.

[TomMarius]

What do you think about WebAssembly?

[patrickmay]

The last time I looked at it, support for Lisp was questionable: https://news.ycombinator.com/item?id=11269736 I need to check again.

[singularity2001]

https://github.com/appcypher/awesome-wasm-langs#scheme

[trcollinson]

As a theoretical experiment it’s interesting. But it would never ever work. You’d literally spend a year or more trying to prepare for that 100 days and I’m still not convinced you’d get anything out of it.

Now, if you were going to give me 6 people that would be a different story. With 6 people I have a number of startup ideas that I would love to work on and I think I could get a working mvp out of in 100 days.

As of today it would be around health care billing for small to medium sized US based medical practices encompassing ehr, billing, office management, and cash flow management. Not exciting but it’s what I’d be doing.

[posix_compliant]

You could make 16 teams of 6 and give each one a different startup idea. You could then create some metric for success and then choose to continue to work on the one with the most success.

[trcollinson]

I could. I’ve run very large development groups. It takes a lot of time and effort to get it right and with 16 groups of 6 it would be the same overhead, if not more. It just wouldn’t be for me, that’s all. I think I’d be much more successful with one strategic team of 6. If you can manage 16 teams working on projects for 100 days, by all means go for it.

[pro_zac]

That's great! You could secretly give some teams the same idea and compare approaches and results afterward.

[arthev]

Agreed. I have no idea how to manage 100 people, nor how to utilize 100 people in any effective way.

I guess I would be hacking the challenge by holding some clumsy event on the first day or two where we sorted into 5 to 20 different teams, and then got to work on a lot of things in parallel. And then people could switch around between teams as interested or as they might use their skills.

[theossuary]

I think it's a fair assumption to say one of those hundred would have the experience to manage the other 99, and allow you to play stakeholder, not project director. That said, hard to imagine getting a lot done in 3 months with a 100 person program.

[trcollinson]

That’s exactly it. It’s more the 100 day part that is restrictive. It takes a lot of time and effort to get 100 people working effectively.

[zhte415]

Here's a book recommendation if you're not sure how to manage these 100 people. You don't have to. Just facilitate them to identify what to work on, then to go and work on it.

Open Space Technology, Harrison Owen

https://www.amazon.com/Open-Space-Technology-Users-Guide/dp/...

A quick but insightful read.

[twunde]

If you're actually interested there are a few companies taking this on such as Athenahealth and DrChrono, both EHRs for small to mid-sized companies. Or if you want to focus on healthcare billing, you should check out Cedar, which focuses on large systems.

[newman8r]

I'd probably have them work on finding as many zero-day vulnerabilities as possible - after the 100 days I'd probably keep the top 5-10 performers on salary if I was able to make enough money from the research.

[neals]

I would take them away from what they have known and done for years and years. I would finally take their skill and ambition seriously. They will work on a goal that is beyond what they could even imagine when they learned and studied and improved their knowledge while others partied and explored sex and drugs .

Bacause for me, they would be figuring out how to have more people click on my ads.

[zaphirplane]

Subtleties are lost in the unwashed masses, too funny.

[agravier]

An Amazon competitor. Either in a few underserved niche markets, or in an area where Amazon is not very present.

[zaphirplane]

Post the question on reddit and HN, wade thru 99% bad humor theads, and borrow the best idea

[PappaPatat]

Build an app that would remind you about things to do when you are in the vicinity.

Rational: there are plenty of things in my live I should do that are not time based, but location based. Post a package to a friend: when close to a postoffice. Buy shaving foam, when close to a supermarket. Buy these earrings, when close to that jewelry shop where she pointed them out to you. Buy flowers, when close to a flower shop. See that exhibition, when walking by. Visit a friend, when close.

Include the option for businesses to send you sweet deals when close and you'd be making money to pay the salaries.

[scotty79]

I would make them read everything there is to know on superconductivity (esp. high temperature superconductivity) and make them brainstorm ideas on how to get us closer to room temperature superconductors.

[itomato]

100 people @ 100 days / 6 hours per day is 6,000 person-hours at best.

30% of that time or more would be spent on consensus building and chatter.

10% on LaCroix & Coffee runs.

5% on bio-breaks.

I'm left with about 60% of the original time, which I would have to spend on the problem of preserving the original allocation.

Certainly some software product would be the result, negating all the ritualized product management activities we hold dear.

It would require ego death for all involved, as well as the willingness to cast aside "reason" for the sake of gain.

[miguelrochefort]

60,000 person-hours

[znpy]

I'd probably work on a mobile phone, based on some sort of SBC like rPI Zero or the Chip Computer.

But I'd work more on the hardware than on the software: I want a physical keyboard like the one the Nokia N900 has and I want a physical kill-switch for the GSM/UMTS/whatever radio.

As for OS an apps, i always thought that some GNU/Linux distro with some small-device-optimized wm and good selinux per-app policies could work very very well.

[slimshady94]

A decentralized ecosystem for government and finances. It would be a good fit as the domain is B2B, working and partnering with governments and banking institutions (with a focus towards developing/third-world countries to jumpstart their standards of living). The ICO would ensure their salaries later.

[user7878]

- Schedule a Hackathone of a single person for First 2 days on 10 different domains that not exists.

- We will have 100 MVP / POC after that (10 per domain)

- Create 2 teams in each domain to develop 2 product for each domain out of all best thing from MVP

- We will have 20 product in which 10 will surely market leader

[pro_zac]

I recently watched a video of Kevin Horton discussing creation of the Super NT, SNES on an FPGA. He mentions doing the same for the N64 would be extremely difficult. It's not gonna change the world, but it would be a fun challenge to throw some resources at.

[itomato]

Youre entitled to your own wishes, but really? Out of anything at all, you'd put 100 (talented) people to work for 100 days to implement an obsolete console in FPGA?

Think big.

[harlanji]

Have them turn on my youtube engineering vlog and work on something interesting. Self organize onto teams using a compatible approach. I’ve just a few months of lectures recorded, but I’d have them give me enough money to keep supplying them with vision.

[adiusmus]

Figure out what to do with people with IQ less than 90 that doesn’t involve turning them into landfill. Seriously. We have plenty of people who aren’t that academic and the world is increasingly leaving them behind. How do we usefully deploy them?

[miguelrochefort]

We turn them into artificial artificial intelligence.

[jacknews]

Exactly. There's an article on HN today describing some AI startups secretly using mechanical turk etc rather than actual AI.

[minikomi]

Real life Battle Royale match broadcast on pay per view. Solves the salary problem nicely.

[oceanghost]

Build educational tools. Bring the cost of learning as low as possible.

[gorbachev]

Figure out how to convince Jeff Bezos to use his money on endeavors that improve the lives of people on this planet rather than billionaire vanity projects like space rockets or newspapers.

[potta_coffee]

I'd split them into 20 five-person teams have them each come up with a different prototype based on whatever best idea they can come up with as a group.

[xellisx]

Probably something (or things) along the lines with helping out with the issues that low income families face, and more specifically the children and their education.

[bsvalley]

I'd tell these 100 Amazon employees to finally go on vacation... and for a 100 days :) Let's not burn these "resources" out.

[analognoise]

I'd fix kicad.

[rartin]

I would fix the sad state of non-profit/charity management; which is currently inefficient and expensive.

[worldsayshi]

I'd have them try to build a tool that allows finding the consensus answer to this question.

[phendrenad2]

I would split them into small teams and call it a consultancy.

[crb002]

Affordable fresh meal kits in urban food deserts that other retailers can franchise from from Amazon/Wholefoods.

[purplezooey]

Gerrymandering

[gameswithgo]

tw2002 mmo

[DrNuke]

I would train them on liberal arts, philosophy & politics so that we have 100 techno-humanists able to connect more dots while thinking of the future of mankind.

[matte_black]

Assuming everyone knows each other and is ready to work, I would introduce myself and my idea: A dating social network that uses genetics data as one of the criteria for matching people.

I will select the top managers in the group, assign each a responsibility, and then have them go into the crowd and build out their team to help accomplish that. I would have the best pitchmen go out and try to raise money from investors.

Everyone goes to work while I sit back in my chair and tent my fingers.

[berbec]

<reductio ad hitlerum>Didn't they try that in 1930s Germany?</reductio ad hitlerum>

[newman8r]

Are you actually working on that? Sounds cool.

[calciphus]

Sounds like eugenics

[theeeerob]

Setting aside the fact that more diverse genes generally result in healthier offspring...

[newman8r]

Not necessarily. It could just take a lot of data on compatibility and see if it could correlate it to genetics. Whether or not there is any correlation would be the question though.

[matte_black]

Nothing wrong with letting people find the most optimal mate with which to produce offspring.