Ahh good memories. I wrote Pente for PalmOS back in 2004 or 2005, complete with a computer opponent using a minimax tree with alpha/beta pruning. Unfortunately I never finished or released it. It was so much fun to do early mobile development - it compared to nothing else I'd done at the time. And it wasn't easy. This is definitely an accomplishment.
Also big fan of vagrant global-status. I have it aliased to vgs in my terminal.
Also very helpful for those using Vagrant on OSX is Vagrant Manager [0] which gives you menu-bar integration and a quick interface to turn VMs on and off. It's useful even to remind myself when I've left some VMs on, especially if I'm running on battery.
Some food for thought: "The future of employment: How susceptible are jobs to computerisation?"[1] which has been discussed on HN previously, lists "Appraisers and Assessors of Real Estate" at .9 probability of computerization and "Real Estate Brokers" at .97. It's certainly not a solved problem, but these two jobs are nearly at the top of the list (or bottom, as the list is in reverse order).
Zillow and Redfin have attempted to automate just the valuation of properties (with enourmous resources behind them), and still aren't as accurate as traditional appraisers. It's easy?
Edit: Even Zillow agrees. There's a million dollar prize if you can do better.
So? Real estate agents aren't appraisers and shouldn't do "official" appraisals because of misaligned incentives (among other reasons).
Appraisals are kind of weird anyways, they are only an somewhat educated guess until someone actually buys the property and they are pretty individually subjective. My friend had 3 different appraisers come to her house and do three different appraisals and they all varied by quite a bit and one was absurdly high, much higher than the others, and extremely unrealistic (IMO). Appraisers also have information that algorithms don't because they actually go to the house and view the property with their own eyes. They can take into account better the condition of the property and "soft," more subjective inputs that you can only tell by actually viewing the property. It's not really directly comparable to an algorithm since there's just a subjective bend to the whole process and there's much more information available to appraisers.
Not only that but when the bank does an appraisal for a mortgage the appraiser's job is only to justify the selling price. They start with the agreed upon selling price and they use data to justify it, they don't go into it blind, they consider and analyze the agreement of sale as part of the valuation. Zillow obviously can't take that info account as Zillow doesn't have that information. So you're comparing apples (a market analysis based on public information) to oranges (a justification of a privately agreed upon sale price).
An appraisal is also cheap, especially compared to a real estate agent. Ours was a flat $425.
Fair point! A realtor can still walk a property to perform a market analysis, which is still more accurate than an automated valuation.
I'm exhausted by the never ending comments here "it's easy! Anyone can do this!" when it's clearly not. I'm simply countering the disingenuity, whether the source is ignorance or arrogance.
In markets where anything can happen and buyers overbid on the order of 100-300k over asking price simply because they want it and will do anything to get it, it is very difficult to accurately estimate the value and/or selling price of a home.
also, ever wonder why the home appraisal comes in right at the offer price during the loan process? its not a coincidence. the appraised price typically has more to do with conducting a business transaction in the interest of all parties involved and less to do with an actual valuation of the home.
The market value of a thing is what someone is willing to pay for it. The appraisal usually comes in at the offer amount because the appraisal is ordered by the bank providing the loan. The appraiser knows/has the offer and works backwards to ensure that other people would likely offer the same amount as the current offer/buyer.
Thats what I'm saying. It's rare at that late stage of the process (at least according to my real estate agent friend) for the appraisal to come in off the mark. Nobody wants to see the deal fall through in the last stages.
I agree with you! That's why you can't automate a valuation, unless you can magically poll through digital means potential buyers in a market on demand. You don't know what the property is worth until it's on the market and you have bona fide offers in hand.
>A realtor can still walk a property to perform a market analysis, which is still more accurate than an automated valuation
Wow, how on earth do you know that!? That's a pretty bold statement and I don't really see anything to justify it at all. You start with "algorithm aren't as accurate as traditional appraisers" and then conclude from that "real estate agents (who aren't appraisers!) can give a more accurate market analysis than algorithms." There's a big leap there.
Even if this was true that real estate agents do a better job than automatic algorithms it's still kind of irrelevant because an appraiser cost somewhere in the neighborhood of $300-$600 where as an agent costs $5,000-$10,000+. If you need an appraisal you hire an appraiser, not an agent.
Don't professional appraisers actually go to the property and look around?
As for the rest, why should the process of selling a house require "expertise"? Apart from inspections, which require domain knowledge of infrastructure.
>why should the process of selling a house require "expertise"?
At a fundamental level, one could say sellers don't require outside expertise to sell a home. Real estate agents are optional. There is a concept called "for sale by owner" which is a strategy some homeowners use. They can list the house on their own (not on MLS though), vet potential buyers, coordinate the showing of the home (appointments and/or open houses), etc. They can also get a real-estate lawyer to draw up contract papers, handle earnest money, etc.
However, most sellers do not want to mess with all the above. They literally do not have the "expertise" to do it all themselves. (Yes, the homeowners can sell a car on Craigslist but selling a home is a step change in complexity.) Hence, you end up with middlemen like real estate agents.
Some people think real-estate agents only exist because they have a monopoly on the MLS computer listings. That's not true. Even if somebody disrupted MLS database with Zillow listings, real estate agents would still exist. Selling a house is a complicated enough affair that intermediaries who specialize in it would always be a natural emergent phenomenon. The real economic question is whether they can keep charging 6% commissions as the internet evolves.
The real economic question is whether they can keep charging 6% commissions as the internet evolves
That's the point. In pre-internet ages, there was a huge information imbalance that kind of validated that 6% (also, there wasn't such a ridiculous housing bubble).
Someone mentioned a real estate agent making 1250 dollars/hour. I would be willing to jump whatever hoops were needed to make that kind of money if I were in the U.S.
Well, the percentage is only part of the point. Some commenters in this thread are truly perplexed why the agents even _exist_.
They exist because sellers want them to exist. I'm saying it's naive to think that a new revolutionary website with a slick UI/UX would make agents obsolete. Agents may make less money (e.g. forces of free market lowers their median income from $45k to $20k) but the agents won't go away. If you have a population of people that don't want to do something (hassle work of selling) and a segment of population willing to specialize in relieving that hassle, then boom, you inevitably end up with agents. A new website or smartphone app doesn't remove the desires for that business relationship.
>Someone mentioned a real estate agent making 1250 dollars/hour.
Most agents don't make much money[1]. (Median is ~$45k.) There are a handful of superstar agents selling multi-million dollar mansions but most agents are selling more modest properties.
Getting the appraisal from the place that's selling to you seems like an inherent and obvious conflict of interest. Ideally the appraisal should be a flat flee from a third party.
In the context of your comment, what value do you see realtors providing w/r/t appraisals?
Really interesting idea of running separate instances and synchronizing them. I wonder if this could be done in legacy games to support VR - an instance for each eye - and each instance's game state and camera position synced except for slight horizontal offset for left eye vs right eye. Xwing VR!
You'd need to be able to offset the camera position per-instance, which I'm not sure the old X-Wing games can do. I mostly played Alliance, which I'm pretty certain cannot; it has several camera options, but I don't recall any that allow for arbitrary positioning. On the other hand, if you're interacting directly with the game's memory space rather than just its UI, that may not actually be an issue.
Assuming you can't reposition a camera, you could try offsetting the ship position on one of the instances, but then you'd probably have issues with the simulation states diverging - I suspect a big part of why it can be made to work with Flight Simulator is that the state of the aircraft itself is more or less the only thing that's actually being simulated, so you can manipulate it to a fare-thee-well without affecting anything else. Contrariwise, if TIE fighter shots hit the ship in your left-eye instance, but miss the ship in your right-, that'a a much harder problem to solve.
Half of my Amazon packages are now delivered by an Amazon van or by some guy's personal vehicle with an Amazon magnet on it. I believe they're a mix of the Flex[1] program the author referred to and private companies Amazon is contracting[2]. Not a stretch to see them adding pickup as a service and competing directly with USPS, UPS, FedEx etc. See also the companies they're acquiring in this area[3].
I for a time worked for the USPS- what lots of people don't seem to realize is that pickups and last mile deliveries are precisely the businesses that UPS and FedEx want no part of. They are extraordinarily complex in most areas and very expensive. The money is in flying and trucking the packages across the country. They actually pay USPS to do it for them in a lot of places because it's cheaper than doing it themselves.
After all the USPS has been in that business for 200 years and not too many people are bullish on them. Why would Amazon getting into that business be good for them?
