nwenzel

Yes please... Anything to get past crappy mobile websites. At least most mobile websites with apps (or is it baked into iOS?) are smart enough to give you an option to open an app after the page loads.

And from a retailer's standpoint, how could you not use this? You basically bring someone exactly to where they asked to go and exactly to where you want them to go.

Would be interesting to see more GPU equipped machines designed for data processing instead of just processing graphics. ("just" not to minimize the value/importance of design, photography, and art, but "just" in that there's more to GPUs that design, photography, and art.) Form + function for matrix multiplication.

Though totally unrelated to the Mac Pro (other than it being a knockoff of MacBook Air) Dell has made an interesting push for data crunchers. Their recent-ish re-launch of an Ubuntu machine [0] that enables you to simulate and then push your environment up to the cloud is an interesting, if fairly niche product.

Could the Mac Pro, with it's specific audience, the Dell XPS, and maybe even the RasberryPi, be the beginning of a much more specialized desk/laptop computing movement in the face of iPads (errr... tablets) eating the general purpose, internet surfing computer market?

[0] http://www.dell.com/us/business/p/xps-13-linux/pd

While I doubt there will be some Reverse Big Bang of creative destruction, it's tough to argue that the need for knowledge workers is being reduced. In theory, even Excel reduces the need for knowledge workers by allowing a single analyst to produce more output that an analyst with graph paper.

My question to HN-ers, as people who write code that further reduces the need for knowledge workers, how should we feel about our role in contributing to a Vonnegut-style dystopia?

Indifference? If not us, someone else will do it. And job loss isn't really a net loss. It's just capital reallocating to another area.

Pride? Automation brings advances down in price point creating a better standard of living for all.

Something else?

edit: The "need" for knowledge workers isn't reduced. I stated that incorrectly. I meant that fewer knowledge workers are needed for a given task or given output. Clearly, the need for knowledge workers as a pct of the workforce is higher than in the past and will continue to do so until the machines take over.

From the acquirer side:

Maybe they wanted to put some excess cash to work.

Cash is important when you don't have it. But once you have it, it's a dead asset that produces no meaningful return. Excess cash is important as an option on future opportunity. Apparently, Yahoo found their opportunity.

Or, maybe they didn't want to revise EPS estimates down.

An extra $1B worth of $YHOO out in the marketplace would dilutes EPS by about 3% until the revenue starts flowing.

From the acquiree side:

Opportunity vs sure thing + choice of opportunities.

$1.1B in $YAHOO today will be worth something else tomorrow. Maybe more. Maybe less. Maybe $0. Cash is guaranteed to go down in value at the rate of inflation. But not more. And they get to decide on some other opportunity(ies) to go after.

Great post. I think one key to the whole program are the weekly deadlines. And of course the big deadlines. It's the same as not having a boatload of startup cash. It forces you to figure out what is really important.

The trick is, deadline only work if there are consequences. There can only be consequences if you feel accountable to someone or something. So, tell someone you're going to accomplish something by a certain date.

I'm going to contact 25 potential customers next week.

Thanks again Jason. Your advice and insights were very helpful.

I feel obligated to pass on what I learned:

1) Don't forget what it's like to not have your background in your industry.

2) Be very good at explaining what you do. Succinctly. Clearly. And be able to give that explanation without assuming prior industry knowledge (see point 1).

3) If it takes a long time to explain who your customer is, you're doing it wrong.

4) Don't forget that YC partners are also looking for your potential on demo day.

Thanks. Any chance you want to mock interview me?

We build software that stops law firms from overcharging our customers.

nwenzel edgesolutions.com

Email sent with phone number. Thanks! Can't say how much I appreciate it. But, I'll try. Pumped for Sunday! Pumped for phone call today!

Thanks for the post. Interviewing on Sunday.

I've been trying to figure out how much time I might be able to get to explain what we do before the first interruption. 15 seconds? 10? Basically, I get a verbal twitter post, right?

I do. Sample size = maybe 3 (I've moved around a bit). I'm in the US, have lived in Southern California (Edison Electric), Phoenix (SRP?), San Francisco (PG&E), and Boston (NStar pronounced N-Staaahhh by my Mass. friends).

I thought all utilities were on a tiered pricing model. It's actually a technique employed by Arcata, CA to fight pot growing houses [1][2] that use > 600% of the baseline usage.

Also, I have lived places where middle of the day pricing is higher. My utility company just sent me an offer for tiered pricing by time of day. Smart meters help with implementing time of day pricing.

[1] http://www.arcataeye.com/2012/08/measure-i-the-grow-house-el...

[2] http://www.usatoday.com/story/news/nation/2012/11/03/pot-tow...

Just don't forget that you're dealing with problems caused by growth and demand. You're basically being crushed by latent, pent up demand. I feel your pain. But I have to say, having problems for those reasons can't be all bad.

And, if high quality photos are your issue, talk to Pete Warden over at Jetpac. He's wicked smart, friendly, in SF, and deals with high quality travel photos on Jetpac.

Or (And?), talk to AirBnB. They deal with photos, are in SF, and are also YC alums. Don't you all have each other on speed dial?

As a founder, I wish I had your problems. But, I have to say, I also wish I had your level of access to potential solutions and advisors.

My YC interview is on Sunday. Hopefully by Monday we'll have each other on speed dial. :)

OpenShift from Red Hat is often cited as a competitive offering in Heroku vs X reviews. When I first checked it out they hadn't yet come out of beta.

Uses git just like Heroku. Has a similar "dyno" concept but calls them gears. Has a free developer version.

Don't know about support today, but the people I talked to from the team were incredibly helpful.

On kWh costs, don't forget that when you reduce use, you reduce the last kWh first. Meaning, in a tiered pricing model, you reduce your most expensive electricity first.

So, the price you would use to calculate savings is not your average price / kWh. You should use your max price / kWh.

That's also why if/when you install solar panels on your roof, you recover your investment faster if you don't quite cover your entire electrical needs.

Actually, that's not entirely true. If lawyers bill inappropriate charges, you don't have to pay those charges. Inappropriate charges include excessive hours, administrative tasks, overhead, or other charges you disallow based on your agreement (commonly called Billing Guidelines). Based on data I've worked with (insurance companies, financial services company, real estate firms) those inappropriate charges are anywhere from 5-15% of the total. Usually 5-10%.

The advantage of having the VC pay the legal bills is that they are a bigger fish and can more easily put in place and enforce the legal billing guidelines. Also, as a much more likely repeat customer, the VC will likely get better bills to begin with.

Alternative Fee Arrangements (AFA) are becoming more popular in the insurance world. An insurer will hand over substantially all of their plain vanilla cases in a certain line of business (usually workers comp) in a specified geography to an outside counsel firm. That firm is paid a fixed fee per case, not per hour. To protect the firm, the firm is allowed to "opt-out" of some number of claims each year and bill those claims hourly.

Insurers are reporting significant savings. Law firms that get the work are happy. The rest... not so much.

The bad old days are still alive and well in insurance and banking. Bankruptcy is even worse.

Eight years in the insurance world has shown me that the two biggest law firm "bad behaviors" are excessive billing and multi-party meetings. Both of which, as noted, VCs will be in a stronger position to refuse to pay.

I call this version of the principal-agent problem the "new boat problem." When a partner wants a new boat, she/he bills more hours.

Even more pumped for YC interview on Sunday after reading the comments.

Love that this is #1 on HN and I'm going in for my YC interview on Sunday to explain how we can help insurers, banks, and VCs (and apparently startups) reduce their legal expenses.

It's a sign!

For a more substantive comment, moving the legal bills into the hands of a "bigger fish" is bad news for law firms. VCs are less likely to be willing to pay for over-billing, unnecessary tasks, and other bad behavior. (shameless plug) VCs, especially those that use my team's software, can save a bundle on legal fees with very little effort.

Cloud storage companies reduce their costs and storage needs by analyzing stored content to find duplicate content (and I would think partially duplicate content) so that they can store only single copies.

I think BitCasa offers unlimited storage. They had stated some time ago that they can offer unlimited storage because their algorithms for content matching were very effective at limiting what they actually had to store. Size becomes a competitive advantage.

I think the OP and my own experience with Instacart (see below) highlights the importance of high-quality customer-facing employees for startups. If the CEO has to step in to solve simple problems, you're probably doing it wrong.

We had a weird issue with Instacart. We purchased $83.27 worth of groceries. Two items were out of stock so they refunded $10.99. Except... the two items cost more than $10.99. We've been trying to get the right refund but their customer service reps keep saying we already got the refund. Yeah, the wrong amount.

Why the big fuss about saying some startups are "bad"? I'm sure some startups are bad. There's certainly a possibility that some bad startups looked good and made it through the screening process. Other accepted bad startups may have been risky bets that didn't payoff.

I think at least a few people have a problem with the directness and harshness of the statement. But if I was one of those bad startups, I'd appreciate the candor and directness. That would be far preferable to someone being "nice" and allowing me to continue down the wrong path.

Being "nice" isn't really that nice at all.

People seem to be really put off by the lack of interactivity. Interactivity might be fun, but it mostly gets in the way and can be an impediment to telling a story because it let's developers pass the buck to the end user.

Interactivity is just a feature. What's the real benefit and purpose? So users can navigate the data? OK, then what? So they can find meaning hidden in the data? Code capable of multidimensional analysis is going to be more effective than eyeballing a 2 dimensional representation of the data.

Don't force your users to move the mouse around, hover, click, and search around for meaning.

For all charts, figure out what story you want to tell, then tell that story. Don't force your user to search for meaning.

I saw the same thing. PG is pretty consistent with that tagline. So presumably the VCs knew it would be there. If those VCs didn't want to give the appearance of agreement, they probably would have asked to have their names witheld.

PG also specifically mentions VC "noobs" as being a motivation for this protocol. I would think that experienced VCs would be in favor of such a protocol so that "dishonest" or noob VCs wouldn't disrupt a potential deal or increase the cost of the deal with disingenuous offers.

On a Mac, cmd-1 or cmd-2 or cmd-3 or cmd-n jumps to the 1st, 2nd, 3rd, nth tab.

I found Machine Learning for Hackers to be good as a practical, case study based book for learning to apply and understand what is going on in several machine learning problems.

http://shop.oreilly.com/product/0636920018483.do http://www.amazon.com/Machine-Learning-for-Hackers-ebook/dp/...

It's definitely a practical approach. There's a lot of explanation but, despite coming from two PhD candidates, it certainly did not read like an academic paper.

It's also case study based, not necessarily algorithm based.

As a bonus, I used that book to teach myself R. I don't think it's meant to be an intro to R tutorial, but it worked for me.

All three are positives in my book.

Too bad his message was lost by writing a 1,211 word article, explaining 2 different suspicious reviews, with multiple graphics, each with way too many words to read.

Would have been the perfect time to sit on his hands and hire a PR firm to tell the story. The details may be important in a courtroom, but potential customers just want to feel safe.

Now the story is two guys who don't like each other, screaming at each other.

Great lessons here for office politics. Or the playground.

Liar! You are! No, you are!

As a founder who doesn't consider myself to be the "sales guy", this post changed my whole view. It showed me that sales is a problem to optimize and solve, not a genetically predetermined outcome.

As for the high cost of living, try looking at it from a different angle. Instead of thinking about how much more you can get for your money elsewhere, ask why are all those other places worth so much less.

I'm not saying the math adds up to you needing to move to SF, NY, Boston, or other city. But this post helps explain why SF is valued so highly by the many people who choose to live in here.

Other than the architecture of their multi-cluster management system, which us certainly fascinating, I found two interesting points.

1) Use of S3 as the data storage platform allows for Production, ad hoc, analytic, and "personal" clusters to all work with not only a full-size data set but the same data set. Brilliant.

2) The investigation of Amazon Redshift as a Teradata replacement. First, it validates Amazon as a new competitor to on-premise relational data warehouse implementations. Second, it would seemingly move substantially all their data to the cloud. Third, what does that mean for their other apps such as "traditional" BI? No sense having Cognos on a local server if all the data is in the cloud.

Random Forests are interesting because they are designed to make many, many bad decisions that cancel each other out leaving only the good decisions behind.

To oversimplify, each of the many, many decision trees is going to do a poor job building a classifier. Each tree only uses a subset of the data and a subset of the features. So, the results of each tree won't generalize well (and may not even do well classifying the data it was trained on). But, for each tree that misclassifies in one direction, another tree will misclassify in the other direction. The theory (which appears to be validated in practice) is that poor decisions cancel each other out but good decisions don't get canceled out.

Although, it isn't quite a simple as a "good" tree and a "bad" tree. All decision trees are bad but there should randomly be good parts in the bad trees. Each tree is going to be "mostly crap" (as Jeremy Howard puts it). But "mostly crap" means "a little good." The crap in one tree cancels out the crap in another tree. The good parts are left behind... and... like magic... you have a good classifier that is fast although somewhat of a black box.

Jeremy Howard from Kaggle is a big believer and where most of my knowledge of Decision Tree Ensemble algorithms come from: https://www.kaggle.com/wiki/RandomForests

I can guarantee it will go for more than $500. If not to me for what I bid, then it will go even higher than what I bid... which is a secret. So no one bid $502, okay.