m_ke

60 ms·

691.
by m_ke 7y ago
Income inequality becomes an issue for the rich in cities like SF and NYC when costs of living rise faster than wages for the lower class and you end up with a rising homeless population.
692.
by m_ke 7y ago
I don't have the stats but I'd assume that the cost of living has also gone up in the largest cities along with the density.
693.
by m_ke 7y ago
Listening to boring podcasts or reading machine learning papers in bed really helped me deal with my stress related sleeping issues. I'd be fine all day, but when I tried to sleep my mind would start racing and I'd start stressing
694.
by m_ke 7y ago
I agree with you that on aggregate across most cities all of this holds, but it's not alway the case for top cities like SF and NYC, where the true demand is way higher than potential supply. I'm not a NIMBY type and am not compla
695.
by m_ke 7y ago
The "boom" is relative to all of 2000s. Look at the number of permits in 2015 [1], especially in the outer boroughs. My argument, as someone who grew up in Brooklyn and Queens, is that building new housing doesn't really driv
696.
by m_ke 7y ago
Manhattan has a glut of luxury condos that are sitting vacant but Williamsburg, Bushwick, Greenpoint and LIC have seen a boom ( https://www.wsj.com/articles/a-building-boom-pushes-new-york... ) over the past few years. C
697.
by m_ke 7y ago
Here's a link with some more data https://streeteasy.com/blog/nycs-unsold-condos/ . The rezoning and redevelopment of the queens and brooklyn waterfronts brought a ton of new high rise condos with it. I live i
698.
by m_ke 7y ago
All of the people I know who left NYC also left because of housing costs, even though New York has been putting up new housing at an insane rate. https://www.nerdwallet.com/blog/mortgages/migration-study/
699.
by m_ke 7y ago
Yeah, on the applied side there's still a ton of work left to be done to make productionization of machine learning systems easier. We still lack good tooling for data annotation, experiment tracking, model calibration/evaluation
700.
by m_ke 7y ago
Research progress might slow down for a bit in some areas of machine learning but the commercialization of existing technology will keep us busy for the next 10 years. Unlike the past two winters, deep learning is actually enabling a ton of
701.
Ask HN: Best resources for learning about the business side of startups?
9 points by m_ke 7y ago|2 comments
702.
by m_ke 7y ago
If you want to get into philosophical questions you should get some actual philosophers on the show because right now it seems like most of your guests are just making up answers to your questions on the spot and it's not very insightf
703.
by m_ke 7y ago
Yeah, in the interview with Kahneman he doesn't even really get into any of his work in cognitive psychology and behavioral economics, instead he wastes half the show on hypotheticals about AGI. We're as close to AGI today as we w
704.
by m_ke 7y ago
Yeah I have an issue with that as well. He gets some great guests on the show and spends most of the time pushing a philosophical debate about AGI and futurism instead of letting the guests talk about their areas of expertise. I'd even
705.
by m_ke 7y ago
> You should try to limit yourself to opportunities that could be $10 billion companies if they work (which means they have, at least, a fast-growing market and some sort of pricing power). The power law is that powerful. This is easy t
706.
by m_ke 7y ago
I've struggled with this a lot as well. I think the solution is to work backwards, get automated deployment working first so there's nothing stopping you from shipping, then work from master to force you to ship small changes with
707.
by m_ke 7y ago
Gatsby with mdxjs on GitHub pages built using GitHub actions is a pretty good. A simpler mdxjs based alternative to Gatsby would be even better but I haven't seen one yet.
708.
by m_ke 7y ago
I was just looking to see if there were any decent ereaders that could handle research papers in PDF format. I couldn't find anything but I guess there's hope for one in the future.
709.
by m_ke 7y ago
Currently they're just causing a ton of ear infections and hearing problems. https://www.reddit.com/r/airpods/comments/e2gmyw/ear_infecti...
710.
by m_ke 7y ago
Operations research and business analytics will obviously continue to be important but we'll also start seeing machine learning replacing a lot of labor and augmenting a large portion of our daily tasks. It will change how we interact
711.
by m_ke 7y ago
Sounds to me like an issue that comes with coupling of validation with serialization. A lot of these API frameworks combine the two, with a the goal of automating validation when receiving data from clients, but then also do that validation
712.
by m_ke 7y ago
That's definitely true, maximizing shareholder value is all about extracting as much value as possible from your customers, until it no longer works. I'm against most forms of recommendations as well, that doesn't mean that t
713.
by m_ke 7y ago
I've been doing machine learning for almost 10 years now and can tell you from experience that when a layman hears AI they think AGI, a magic pixie dust that can solve any problem that they throw it at. AI is a branch of computer scien
714.
by m_ke 7y ago
> ML is certainly valuable for speech recognition, translation, and fraud detection, but those are somewhat niche applications. It's a long way from "every tech company, large and small". Those were just some of the most v
715.
by m_ke 7y ago
I'm saying that they have systems in place to measure the impact of the code that they put in production on their bottom line and if those recommendations didn't move their profit margins in the right direction they wouldn't
716.
by m_ke 7y ago
Does it matter? They're a manufacturing powerhouse and have a government in place that can (and has shown willingness to) take long term bets that are in the interest of their country.
717.
by m_ke 7y ago
Recommending something you bought before is probably a much better bet than showing you random items from their huge catalogue. You can have a system that generates them a ton of cash while making mistakes in some cases, outliers are inevit
718.
by m_ke 7y ago
Amazon uses machine learning to forecast demand and ship products to warehouses before you order something. It's part of what makes their 1 day deliveries work. https://www.npr.org/2018/11/21/660168325&#x
719.
by m_ke 7y ago
AI and machine learning have no business being mentioned in the same sentence. The AI hype is pushed by marketers and journalists because it gets clicks and invites the futurist crowd to chime in. We're decades away from anything that
720.
by m_ke 7y ago
Machine Learning is a technology the same way blockchain is. Customers buy products and solutions, not your tech stack so it doesn't make sense for any company to sell themselves as a machine learning company, unless you're sellin

More ›