HNHacker News
TopNewBestAskShowJobs

nishantvyas

159 karma · joined April 1, 2014

I do databases. https://www.linkedin.com/in/nishantvyas/
submissionscomments
nishantvyas··on Waymo’s Backseat Drivers: Confidential Data Reveals Self-Driving Taxi Hurdles
Correct.

It’s amazing, how much forgiving we can be for human errors ( accidents every year) but absolutely not for machine/autonomous vehicles, even when, statistically speaking, machines may make better decision much faster(or at-least no worse than human judgement)... I guess feeling/perception of being in control is more important to us...

Other interesting observation I find in every autonomous vehicle discussion is, how we only focus on edge cases... when in reality every tool that we use (including the car we been driving) today are built for general use case and operate under mostly a control environment.

Rather if we think autonomous car as additional pair of eyes and hands when we need it most might serve us well in short run before the technology gets mature over next decade or two.

I’ll be really happy and relaxed if my car can mostly (70-80%) drive it self to daily commute or next trip to LA; expecting it to be my chauffeur is bit too much, personally.

nishantvyas··on Ask HN: Is this a good fit for a column-oriented database?
Things that would work in favor of Column Store (like Cassandra or MariaDB ColumnStore or Redshift etc),

- Everything stored in one table/collection, with 100+ columns/fields

- Many columns will have numerous duplicate values (often with a small handful of distinct values accounting for most of the data, and a long tail of rare values, e.g. most "favorite ice cream flavor" data clusters around vanilla, chocolate, and strawberry)

- Pre-defined schema, but each item will only have values for a subset of the columns (5 - 20 on average); everything else will be empty / null

- No need for a sophisticated or relational query language, but basic aggregation support is needed (ideally group by count, sum, and average)

- A very common query will be a "group by count" for a few columns

- Most queries will only select a few of the column values

- Support for streaming ingestion is a plus

- There will be about 3-6x more reads than writes

- Potential for data loss is acceptable

- No need for an in-memory datastore

Not so good for Cassandra,

- Regex support required in the query language

- You did not mention if most queries would be accessed by primary key; if not, then this will be an anti-pattern...

nishantvyas··on Ask HN: Recommendations for financial open source databases?
What is "financial" open source databases?
nishantvyas··on HERA – Caling Database Access for 100s of Billions of Queries per Day PayPal
I posted it here, yesterday, https://news.ycombinator.com/item?id=20509984
nishantvyas··on We Changed YugaByte DB Licensing to Apache 2.0
Going to vendor's website always have comparisons that claims, it can do everything that the competitor's can not. What I'd also like to know is limitations. Where it's not a good idea to use a product? This would help me make informed decision and build relationship with vendor based on trust.

Best of luck with new licensing.

nishantvyas··on No limit: AI poker bot is first to beat professionals at multiplayer game
I don't get this... Poker isn't pure mathematical... it has emotions involved (greed, fear, belief, reading others, manipulations (to fool the opponent)... and may be more... and all of these emotions arises differently for different people based on their time, place, their world view, their background and history...)

Are we now saying that a computer can do this all in simulation? if so, it's a great break through in human history.

nishantvyas··on GDPR Enforcement Tracker: List of GDPR fines
Does enforcement changes behavior? I guess the time will tell. But I do expect some insurance companies start selling GDPR coverage policies soon.
nishantvyas··on Sam Altman on Twitter: It's gross seeing so many root against Tesla
Pre-cursor: 1. Don't assume anything 2. Don't take anything personally.

Can you help understand what's wrong in that statement, "In the 1st quarter, we registered one accident for every 2.87 million miles driven in which drivers had Autopilot engaged. For those driving without Autopilot, we registered one accident for every 1.76 million miles driven. By comparison, NHTSA’s most recent data shows that in the United States there is an automobile crash every 436,000 miles."

Answering your questions: 1. Either Tesla drivers are 4x safer than the average driver or their stats aren't comparable with the industry standard measures (1.76m vs 0.43m) and they have no problem with misleadingly comparing the two.

-> Well, Tesla has one of the best NHTSA safety ratings.. so their driver might be actually be 4x safer... or at-least based on 1.76m vs 0.43m number...

-> one can argue that oldest Tesla on road is 8-9 years and the total # teslas on road are less thus they have lower denominator for total miles driven by all teslas / total # tesla vehicles on the road, which help them get to 1.76M but that doesn't change the fact 1) it is 4x safer based on that number and 2) it has one of the best NHTSA safety ratings.

2. Tesla is not interested in supplying a real metric for how likely you are to crash in a Tesla, but instead provide PR numbers (accident per million miles when Autopilot is overwhelmingly used only on the safest roads?)

-> first, Tesla used NHTSA metric here, which is "NHTSA’s most recent data shows that in the United States there is an automobile crash every 436,000 miles."

-> Based on that metric, they provided 2 data points, 1) "we registered one accident for every 2.87 million miles driven in which drivers had Autopilot engaged." and 2) "For those driving without Autopilot, we registered one accident for every 1.76 million miles driven."

-> that is the real metric on how likely a Tesla car (either autopilot or not) could run into accident... now for "supplying a real metric for how likely you are to crash in a Tesla" (emphasis on "you") would be difficult for any car provider...

I'm strictly commenting for the statement you quoted and the 2 questions you posed. I'm not discussing anything like, how they have over-sold (an idea) of auto-pilot, if its truly safe, if its ready and so on.

nishantvyas··on End of Apple
Smart phone industry will soon enter its adolescent and will show the sign of maturity. Also, smart phones became (or sold as) fashion device compared to computing devices (like computers, laptops etc) in previous decade, which were more like utility. Fashion changes every year, utility does not; may be the market is getting to that maturity too.
nishantvyas··on Two creators are not always better than one
"Solo founders are twice as likely to succeed in business as co-founders" by research from Jason Greenberg, PhD ’09, and Ethan Mollick, PhD ’10 and MBA ’04.

:-)

hmmm... co-researchers to research on NO need of co-researchers/founders...

nishantvyas··on Jaywalking: How the car industry outlawed crossing the road
As we live more urban with growing population, solution like bikes, cycles, skate boards, scooters (lime etc) and of-course more people on foot... will emerge... we will see how old rules play out then...
nishantvyas··on Sequoia Capital on startups and the economic downturn (2008)
timing the time is impossible for any market. It's easy to be contrarian but hard to be a contrarian that will be right (on time).

Reality is no one really knows, what will break the camel's back!

nishantvyas··on Solving Tech Addiction Is an Underappreciated Market Opportunity
very famous investor once said, unless your startup has one of "Seven deadly sins"; its hard to be sticky (or successful). So if one is planning around tech addiction, make it about, pride & envy not about how its good for your life overall....
nishantvyas··on The Only Metric That Matters: How many are really using your product? (2017)
it depends. Are you raising the chicken for egg or meat? Egg is about sustainability... meat is about growth (very fast)...
nishantvyas··on How Our Marketing Team Spends Money Each Month
So to summarize, they have "10-Person Marketing Team" spending 93k/month total, with 2 major breakouts, 1. Teammate Expenses (73,225 + 9,402 + 367) = ~83K 2. Everything Else (or Actual Marketing) = ~10k

Why do they need 10 person team for $10k budget (actual money spent to bring customers in)? unless these guys are literally going door to door.

P.S. I'm not marketing professional.

nishantvyas··on Ray Dalio: the economy looks like 1937 and a downturn is coming in two years
Winter is coming?
nishantvyas··on Why the Future of Data Storage Is Still Magnetic Tape
Well my point was not to show flash as superior but to point out the development in that area... every use-case has reason be it flash or Magnetic... author choose to paint Magnetic as better alternative to Flash (in capacity dimension), which is obviously can not be true for all the use-cases... (same for flash) my comment it to high-light that fact...
nishantvyas··on Why the Future of Data Storage Is Still Magnetic Tape
While, Intel Introduces "Ruler" Server SSD Form-Factor.

"Intel on Tuesday introduced its new form-factor for server-class SSDs. The new "ruler" design is based on the in-development Enterprise & Datacenter Storage Form Factor (EDSFF), and is intended to enable server makers to install up to 1 PB of storage into 1U machines while supporting all enterprise-grade features."

ref: https://www.anandtech.com/show/11702/intel-introduces-new-ru...

nishantvyas··on Microsoft threatens to stop hosting Gab unless posts are removed
I guess that's not the problem here.

Problem is combining ignorance with confirmation bias and then multipling it by distribution engines like FB, Youtube, Podcast etc....

unfortunately information is NOT absolute. it's asymmetrical and heavily influenced by our limited world view, ignorance and confirmation biases.... allowing bad agents to take advantage of this fact and then allowing them to use the massive distribution channels like FB etc. is recipe for unrest.

nishantvyas··on 11% of Americans don’t use the internet
Diffusion of Innovation always has "Laggard" All of us have been in different part of spectrum for various innovations... when the innovation/trend is irreversible, everyone eventually converts.... remember rotary phones, VCRs etc

refer: https://en.wikipedia.org/wiki/Diffusion_of_innovations

nishantvyas··on Why We’re Building Flux, a New Data Scripting and Query Language
"I don’t want to live in a world where the best language humans could think of for working with data was invented in the 70’s. I refuse to let that be my reality."

---- re-play the same sentence for c/SQL/ENGLISH ---

I don’t want to live in a world where the best language humans could think of for "communicating" was invented in the (c. 550–1066 CE). I refuse to let that be my reality.

So, I'm starting new language...

sdflakfj lsjkfaldfj sdfkjaslf dflkasjdfk sldfkjaslf laskfdas....

------ P.S.: Any technology is built over time with sedimentary layers... every layer has played key role in where we are today.. I'd not discount any...

nishantvyas··on Best Buy Is Thriving in the Age of Amazon
every-time I go nearby store it's mostly empty... (price/sq.ft. ratio)

I'm also not happy with their price match... I had a BB gift card so went there with Costco headphone deal to price match, which they did with shipping of 3 weeks from order day... If you'd have to guess it... yes... shipping after the Costco deal would expire... I wanted to give benefit of doubt but when they canceled my order after 3 weeks... it kind of felt like a trap...

nishantvyas··on Python has brought computer programming to a vast new audience
W/O going into the war of I'm better than you (Python vs JAVA)... Python is built to make it easy to write good programs... where as JAVA is built to make it hard to write bad programs.... same objective but different thinking/approach to achieve it... hence one(Python) would eventually have larger adoption....
nishantvyas··on Should I Learn Java in 2018
Any fears of oracle stewardship?

https://www.google.com/amp/s/www.infoworld.com/article/32841...

nishantvyas··on Instead of ‘finding your passion,’ try developing it, Stanford scholars say
Problem is, "Find and follow your passion" social expectation from very early in life. In reality, 1) "finding passion" early is hard; Before one turns 12, brain is not fully develop to logically understand the true likes then from 13-19 it goes thru a huge amount of hormonal, emotional and physical changes... what's left is an average 19 year old teen with I'm "not sure my major" in college, let alone the passion.... 2) "follow passion", again the twisted notion where one's work should absolutely be their passion too... otherwise it's just a "hobby" and hobbies are supposedly not helpful....

If one finds their likes early, become really good at it and if that skill pay really well... your are top 0.1% in your field.... for the rest... build a skill that lands you a job then experiment with different hobbies... when you find a hobby you really like hopefully by age 30-35... turn that hobby in to passion, you still have 30-40+ year of average lifetime remaining....

nishantvyas··on Bitcoin’s Price Was Artificially Inflated Last Year, Researchers Say
You also have to account the rise in population, which translates to higher demand. Not sure Japan had similar Demand/Supply to SF before bust.

The average annual gain in SF since 2010 came to 11,173 persons, and the median is 10,824. The overall population increase of 78,593 amounts to an estimated spike of more than 9.75 percent. For comparison, that’s a higher year-over-year increase than seven of the ten most populous counties in the country, including number-one ranked Los Angeles County (up 0.1 percent since 2016), Orange County (0.4 percent), and San Diego County (0.6 percent).

Breakdown of how the city has grown since the last full census in 2010, with SF’s official population of 805,770 at the time:

2011: +10,524 (816,294) 2012: +14,112 (830,406) 2013: +10,486 (841,270) 2014: +11,988 (853,258) 2015: +13,062 (866,320) 2016: +9,783 (876,103) 2017: +8,260 (884,363)

1. https://sf.curbed.com/2018/3/26/17165370/san-francisco-popul...

nishantvyas··on Tech care of elders
I really liked the CareCoach approach for solving, probably, one of the most profound problem of care taking. Good way of giving the human touch to AI. Also intelligently lowering the bar of AI expectations thru an Pet avatar...

I’ve never used or know someone who has any experience using it but reading the article makes me think they must be doing a decent job...

nishantvyas··on Ask HN: Why don't I believe blockchain will change the world?
When I buy/sell bitcoin on Coinbase, am I considered part of network? Yes the transaction will be part of the blockchain but I am not participating directly rather using proxy (in this case, coinbase for it). This is what I meant by powerbroker.

Second, to my scaling point, In order to be a part of a network one has to maintain a complete blockchain on local server/computer. This is something just not scalable in practice for 99% of people or the use-case... so regardless of the intent there will be an emergence of powerbrokers....

So using the internet analogy, yes internet is decentralized but who owns the backbone (att and few major players) will be the case for blockchain.

nishantvyas··on Distribution vs. Innovation
Yes. when building product (more so for consumer product) there's only two important rules, how to people find you when they are looking for the value your service/product is creating (DISTRIBUTION) and how would they know if they're currently not looking for it? (MARKETING)... everything else... features, technology etc can be solved one way or another...
nishantvyas··on Ask HN: Why don't I believe blockchain will change the world?
who will mange the per use-case blockchains (the distributed ledger for a given use-case)? I like decentralization... more power-to-all...but the reality is eventually there will be power-brokers (for example, miners in bitcoins or exchanges like coinbase) who has deep pockets to manage and scale these blockchains.... there will always be ripples and ethereums built blockchains which mostly maintained by single player... but then how is it any different than now... what am I missing?
← PreviousPage 2 of 3Next →