HNHacker News
TopNewBestAskShowJobs

mhneu

1,420 karma · joined August 25, 2016

submissionscomments
mhneu··on Monsanto ordered to pay $289M damages in Roundup cancer trial
Source: “The Chickenshit Club” by Eisinger. Explains how DOJ and SEC have moved to civil actions because they’re easier to prosecute, letting individuals at corps off the hook for actions they’d be criminally liable for if they took them against an individual in their individual capacity.
mhneu··on Monsanto ordered to pay $289M damages in Roundup cancer trial
But corporate culture can matter. And culture affects who gets hired and how the employees act, even when new people come in.

If you ask biz school profs, culture is essential.

To me the bigger problem about treating corps as people is corps are willing to renege on any deal if it benefits them, and pursue profit at all costs. If a person you know acted like that, you’d call them a sociopath.

So “Corporations are people: sociopathic people.” Is probably the best description.

mhneu··on Monsanto ordered to pay $289M damages in Roundup cancer trial
Do you think they’re paid?
mhneu··on Stop Validating and Start Co-Creating
Summary: don’t let the perfect be the enemy of the good. And don’t overoptimize before you build something.

Pretty good advice for any creative engineering discipline.

The hard part is always defining “overengineered”. Is 1 day of design work enough? Is 2 weeks too much? Good taste and experience rule here, and the best engineers make the right decisions, leading to great products. Problem is, it’s hard to define what makes someone good at this- though it’s incredibly important for product success.

Software design is still a very creative endeavor and probably always will be.

mhneu··on Worst Computer Bugs in History: Therac-25 (2017)
I think unfortunately this is going to change with the advent of "AI" and related technologies, such as autonomous driving

Yes, the potential cost of software bugs is increasing as software does things that no hardware interlock can stop. And worse than that, as a society we largely haven’t realized we need to optimize for worst case (not average) performance of algorithms, because they WILL be attacked. If you’re lucky, they won’t be attacked by sophisticated, well resourced nation-state attackers. But sometimes that will happen.

The rise of complex algorithms to control complex processes is the real difficulty. Facebook’s banning algorithm is an example of something that has been exploited by attackers.

Let’s hope voting software is not the next target where bugs can be exploited. Because changing political decisions can and does produce life-changing effects.

mhneu··on How I gained commit access to Homebrew in 30 minutes
All three responses 'time is money'.

Yes, security is a cost. There's a bit of tragedy of the commons effect here - many of the downsides are pushed onto others. I like Doctorow's general take on socializing costs of privacy and security breaches while privatizing profit: https://locusmag.com/2018/07/cory-doctorow-zucks-empire-of-o...

mhneu··on Engineers teach a drone to herd birds away from airports autonomously
Yes, agreed. What we've learned over the past few years that where algorithms exist, there are incentives to try to exploit them. Ad and PR firms: Google search algorithms, Twitter ranking algorithms. Political orgs: Youtube ranking algorithms, ad placement algorithms. Repressive govts: Facebook abuse reporting algorithms (done in Ukraine), Twitter hashtag algorithms, tweet ranking algorithms (bots pushing particular political tweets).

Humans may be in some ways easier to influence, but once an algorithm gets big enough, it can be worth attacking if the impact would be huge.

mhneu··on How I gained commit access to Homebrew in 30 minutes
Time is money. Takes work to set up. Resources are being directed at development.
mhneu··on Engineers teach a drone to herd birds away from airports autonomously
Problem is downside risk: drones get hacked, or someone figures out how to make the drones fly into planes, or herd birds TOWARD planes.

As a society we've been pretty cavalier about algorithmic control. Which can work well for a lot of things. But probably shouldn't be trusted (without a TON of validation and regulation, as we've done over decades with airliners) with controlling stuff that can injure people or damage societies.

mhneu··on Traceability on the Internet
Yes, but note from the page: those unauthorized accesses were reported and the employees were sanctioned including demotion, etc.

On the private side, it can be worse, because the governance structure is not always there.

https://bgr.com/2016/12/13/uber-privacy-and-security/

Uber employees helped ex-boyfriends stalk ex-girlfriends, and were even able to access trip information for celebrities like Beyoncé, Reveal News explains. These revelations come from the company’s former in-house forensic investigator Ward Spangenberg.

Agreed from comment above this one - it's a problem both for govts and for private industry. I think what democratic societies need is an agreed framework for how as a society we regulate such data collection (as Schneier argues in the link I posted above, or as Doctorow has recently argued.)

mhneu··on Traceability on the Internet
What we've learned over the past year or two is that everything we all used to worry about NSA doing, Facebook is ACTUALLY doing.

I was also one of those misled, and now know the NSA has controls and a democratic governance structure atop them (as do the other four agencies in UK, Canada, NZ, Aus) that restrict what they do for lawful, democratically-endorsed uses only.

Surveillance capitalism at Facebook (good description here: https://www.schneier.com/blog/archives/2018/03/facebook_and_... ) is a bigger threat to democracies.

mhneu··on How to explain infinity to kids
A professor of mine once said to me that all teaching was a process of lying, and then of replacing the lies with successively better approximations of the truth.

This is the opposite of what Feynman thought. He said, and I agree, something like "the hard part about teaching is making concepts simple without saying things that are false." The whole challenge of teaching is NOT lying, but instead saying simple things that are still true.

That said, I see the appeal of choosing omega.

mhneu··on Bay Area cities are cracking down on free food at tech companies
Megan McArdle has an agenda. I've never found she generates much insight for that reason. I apply the Two Papers Rule to her, and to B Stephens:

https://twitter.com/Noahpinion/status/1017590957984735232

There are some worth reading, like Will Wilkinson, J Rubin, and T Nichols. But they're few and far between.

mhneu··on Tesla is building its own AI chips for self-driving cars
A thoughtful point.
mhneu··on Three Levels of Wealth
This is a great comment.

the lack of burden of worrying about how I would pay for X, Y, or Z

I had the same experience of once worrying about that, and now not really having to. It's incredible. Imagine what it's like raising a family while having to worry every time you buy diapers or milk.

And 57% of American adults have less than $1,000 and are in that category of worry. Wealth inequality is the central problem of our time.

mhneu··on Tesla is building its own AI chips for self-driving cars
YES. Security compromise is the huge achilles heel of self-driving cars - and anything that is algorithmically controlled.

When you have 100,000 self-driving cars on the road, and someone figures out how to hack them, and drive them simultaneously into crowds, that will be a BIG BIG problem.

And this problem isn't just about self-driving cars. It is a fundamental issue with all algorithms. It's a lot harder to hack humans to do bad things (although it can be done with sustained messaging and propaganda). But algorithms can be compromised and exploited.

mhneu··on How to tell the difference between persuasion and manipulation
It can be very hard to distinguish these two concepts.

Instead, it is more useful to understand the agenda of the person doing persuasion. Do their interests align with yours? What are they trying to achieve? That's a good way to get insight into whether you're being 'manipulated' to do things that are not in your interest.

Because the agenda of an information source is key, that makes it vitally important to understand who owns and manages the outlets you get information from. In today's world, because of changes in news profitability and thus the number of financially-independent news outlets, knowing the agendas of ownership, management, and funders of news sources is absolutely critical.

mhneu··on Bay Area cities are cracking down on free food at tech companies
The issue is that America funds cities poorly. There is a federal subsidy from cities that goes to rural areas. Red areas get more back in services than they pay in taxes, and blue areas pay out more in taxes than they get back.

So cities have to raise funding somehow.

(This is also a problem with the way America funds schools. The US needs to figure out how to distribute tax dollars more equally - or at least figure out how to keep most of the country out of poverty while still building infrastructure that keeps city housing costs down.)

mhneu··on Confessions of an Ex-Prosecutor (2016)
Probably many people know this, but since it's important to know the owners of our media and information sources:

Reason is published by the Reason Foundation which is funded by several right-wing donors, including the Koch Family Foundations. The Reason Foundation is a member of the State Policy Network whose goal (as measured by their past history) is to advance the interests of right-wing wealthy donors. And leaders at the Reason Foundation have strong ties to ALEC, an organization that coordinates and advances bills across states that benefit right-wing wealthy donors. https://www.sourcewatch.org/index.php/Reason_Foundation

So keep Reason's editorial bent in mind when you read its articles. Some are good, but Reason leadership has an agenda.

mhneu··on Darpa Plans a Major Remake of U.S. Electronics
Why does DoD want generality?
mhneu··on EPA's proposed new rule would allow it to ignore the best available science
Just the authoritarian “conservative” movement that seeks to concentrate power.

I am actually a conservative on many issues, including UBI, work rules, tariffs, caution when making new law and regulations, and appropriate pricing of risk. However the real conservative movement has been redirected as a way to link wealthy donors’ interests with identity politics, for votes. Part and parcel of that is a huge propaganda machine which spouts lies. That propaganda is corrupting the West and I am afraid for its effect on democracy and average people.

mhneu··on EPA's proposed new rule would allow it to ignore the best available science
It's worrying.

The groundwork was laid in the US starting in the 1970s. With the founding of Cato, Searle, Olin fnds. Murdoch bought the NY Post in 1976. Fairness doctrine was repealed in 1987 allowing Rush Limbaugh to go bigtime. Fox started in 1997.

mhneu··on EPA's proposed new rule would allow it to ignore the best available science
From the article:

EPA is proposing to eliminate consideration of scientific studies unless the raw study data are made publicly available — possibly including participants’ personal, confidential, and private information — data that may be subject to legal, ethical, and human subject research protections.

This requirement would eliminate human health studies that use integral medical, lifestyle, and geographic data, as well as studies that include confidential business information, such as information from studies conducted by industry to demonstrate safety of a pesticide or other toxic chemical. It could also eliminate consideration of older studies for which data are no longer available or accessible, even if the data have been reanalyzed and the studies have been validated, replicated, reproduced, and undergone rigorous and independent peer review.

Far from using the best available science for EPA decision-making, the proposal would severely limit the studies and scientific evidence the agency would use to fulfill is statutory mission to protect human health and the environment. This is not a new idea. It is the result of a decades-long campaign to undercut reliance on groundbreaking research...

mhneu··on EPA's proposed new rule would allow it to ignore the best available science
Is it really _Trump's_ EPA that is the problem? Or is the underlying problem the party that backs him and supplies EPA officials?

https://www.vox.com/science-and-health/2017/3/10/14871696/sc...

Scott Pruitt denies basic climate science. But most of the outrage is missing the point. ... It’s not about Pruitt and it’s not about facts. (from 2017)

[Complaints] are all about what Pruitt believes. And in the end, who cares what he believes? He is a functionary, chosen in part to dismantle EPA regulations on greenhouse gases. If it weren’t him, it would be some other functionary.

The goal is to block or reverse any policy that would negatively affect [conservative] donors and supporters, who are drawn disproportionately from carbon-intensive industries and regions. That is the North Star — to protect those constituencies. That means, effectively, blocking any efficacious climate policy (which, almost by definition, will diminish fossil fuels)

mhneu··on People Aren’t Dumb, the World Is Hard
Cory Doctorow has recently written an interesting piece on corruption: https://boingboing.net/2018/07/02/low-yield-crude.html

It talks about how socializing risk and privatizing profits is actually the definition of corruption.

mhneu··on People Aren’t Dumb, the World Is Hard
This is why representative democracies and rule of law are so important.

It's important to make governments that are "about laws - not about men". Because if governments are built around single people, those people will concentrate power. When a system leads to the person in power hands off power to the next person, it's a way to distribute power.

The second principle that is a huge problem today is concentration of power into the hands of the wealthy. Thomas Piketty put his finger on the problem - in the 1970s, most obviously in the US, billionaires began to figure out how to use their wealth for persuasion. In that way they use their wealth to take more of the profits in the economy for themselves.

I think this is the problem of our time: How do we stop oligarchs and billionaires from funding a propaganda machine designed to break democracies so they can accumulate more wealth?

mhneu··on A browser extension to make Medium more readable
Surveillance capitalism. https://plus.google.com/+GregLinden/posts/TWiGesiztbj

(I'm not sure exactly if this is the present motivation. But it's the motivation for a lot of apps. Data can be sold / used for ads when collected in very large numbers, and apps collect more data than websites. So there are strong business incentives to move to an app, for many sites.)

mhneu··on Ask HN: What are your favorite statistics and probability textbooks?
No no on measure theory. See below on Feller, Bersekas, Gallager. I think it's best to learn discrete prob first and skip the difficult math until you decide to become a mathematician or really need it for e.g. finance work.

One other objection: yes, those are fields in statistics. But statistics as a field is MUCH broader than departments called statistics. A ton of stats is done in engineering, psychology, and economics (econometrics). In fact I'd say the major research in stats for the past few decades has been done outside of "statistics" as a field. But yes, the books by Tukey, Tufte, and Casella are good.

If the OP is asking about statistics and probability needed for machine learning, he wants to focus on engineering statistics, like: estimation, stochastic processes, and filtering and mathematical statistics. The Wald rec is good and complements Feller.

mhneu··on Ask HN: What are your favorite statistics and probability textbooks?
Yeah, I personally don't love Papoulis. But it has been regarded as the standard stochastic processes text.
mhneu··on Ask HN: What are your favorite statistics and probability textbooks?
If you're working through Spivek and Apostol, get Feller v1 and v2. Same level of rigor and emphasis on intuition. (If you're having fun after that, pick up Gallager's book on stochastic processes. Similar approaches, intuition, and focus on discrete probability to skip the metric theory complications of continuous dists.)

Another very good book is Bertsekas and Tsitsiklas, Introduction to Probability. This book will also give you some intuition.

I've heard Grimmitt recommended but I think B&T is better.

You could also start with Papoulis (a stochastic processes classic book, but it does the intro probability too.)

← PreviousPage 3 of 12Next →