HNHacker News
TopNewBestAskShowJobs

CogentHedgehog

413 karma · joined May 21, 2020

Software engineer
submissionscomments
CogentHedgehog··on The Secrets of Monkey Island’s Source Code
"You code like a dairy farmer!"

"How appropriate, you code like a cow!"

CogentHedgehog··on The Fake Meat of the Future Is Chicken
I've found that most of the mock-chicken products are pretty good. For some reason, it REALLY IS easy to make things that "taste like chicken." Does anybody know why?

Making a good approximation of beef seems to be quite difficult in comparison -- although the Impossible Burger comes quite close.

CogentHedgehog··on The number of people dying is not the same as in any other year
Yup, the data is right there. HN commenters just want to "but aktsualllly..."

The spike in deaths from the UK's first wave is pretty visible in the graphs, and the spike from cases we're seeing now will probably take a few weeks to show up in the data.

CogentHedgehog··on Unethical human experimentation in the United States
I had no idea how many of these experiments there were.

"Early in the Cold War, in studies known as Project GABRIEL and Project SUNSHINE, researchers in the United States, the United Kingdom, and Australia tried to determine how much nuclear fallout would be required to make the Earth uninhabitable.[90][91] They realized that atmospheric nuclear testing had provided them an opportunity to investigate this. Such tests had dispersed radioactive contamination worldwide, and examination of human bodies could reveal how readily it was taken up and hence how much damage it caused."

I guess that is indeed ONE way to answer the question... the 1950s era had so many absolutely crazy things happen, back when nuclear weapons were brand new and people didn't realize what impacts they would have.

CogentHedgehog··on 1MB Club
Yeah, that sounds about right. I remember MS DOS used to fit on a 720k floppy -- back when they were literally floppy...
CogentHedgehog··on The number of people dying is not the same as in any other year
That's definitely true too -- the rates of mental illness are something like doubled in the base population. But the frequency for people who have been sick is much higher than the general population, even accounting for the spike in overall mental illness.
CogentHedgehog··on The number of people dying is not the same as in any other year
Yes, the long-haulers and COVID-19 heart damage are particularly concerning. Also there is some evidence suggesting that a large fraction hospitalized for COVID-19 end up with mental health issues even if they did not have them before. Some of that might be down to the traumatic experience but there's some reason to think COVID-19 infections cause neurological impacts.
CogentHedgehog··on The number of people dying is not the same as in any other year
Key point:

> The latest figures from the Office for National Statistics (ONS) show the number of deaths registered in England and Wales in the week ending 6 November 2020 was 14.3% higher than the five-year average.

But it is good to know that the measures to control COVID-19 are also suppressing flu and pneumonia (though not enough to make up for COVID-19 mortality).

CogentHedgehog··on 1MB Club
In some ways it's interesting to think that a modern website is larger than most operating systems used to be.

I can understand why (frameworks, rich assets, etc) but it puts us in an interesting place.

I wonder what the web will look like in 2030?

CogentHedgehog··on GrubHub is quietly replacing restaurant phone numbers on Yelp
Grubhub is so shady. Although they're not alone -- the "service fees" some of these food delivery services are charging are highway robbery.
CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
I don't have enough datapoints yet to tell if there's anything suspicious happening around hydrogen tech. The volume of discussion has only started to get high recently.

I'll say that on technical merits there are pretty solid arguments both for and against green hydrogen. It's not like nuclear industry where the free market has already spoken and mostly said "yeah your product sucks, we don't want to buy any more."

CogentHedgehog··on Obama: The internet is “the single biggest threat to our democracy”
This is true. Engagement -> Ad Views -> $$$

If they down-weighted negative engagement they could easily solve this though. There are pretty trivial content analysis tools that spit out a hostility score for a corpus of text.

Facebook just doesn't care to do this.

CogentHedgehog··on Obama: The internet is “the single biggest threat to our democracy”
Hot take: subtle differences in algorithm choices have huge real-world implications. Facebook prioritizes engagement as their top metric. This means that divisive content which inspires strong emotions bubbles to the top of the feed. This has predictable consequences for discussions and communities.

Reddit prioritizes using user votes, which tends to result in less divisive content. But the way the community is broken down into subreddits can result in clear echo chambers at times as people reinforce each other's behaviors.

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
> People shilling for nuclear is an entirely distinct issue, regardless of the data sources they're using to make their claims.

These are people whose behavior shows they are very clearly acting in bad faith -- throwaway accounts that post talking points. They almost exclusively use ElectricityMap to back up their claims.

> Can you source a citation demonstrating "Nobody who is involved with the energy sector thinks ElectricityMap has any credibility."

Of course not, nobody wants to go officially go on the record saying this and get handed a SLAPP lawsuit. But it's a known reality at this point.

I don't know if the issues with ElectricityMap are due to data quality (or lack of data in some cases), but it's not reliable.

Edit: for credible data, go to the official yearly figures reported by national and international agencies.

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
> People popping up out of the woodwork to argue passionately for an out-of-favor technology reeks of a dying industry trying to revive itself with public relations.

Yes. I'll probably get downvoted for saying it, but I want to get out in the open that the situation reeks of a PR/astroturfing campaign. There are a small number of ardent nuclear energy supporters acting in good faith, but a much larger army of paid trolls that have been active for the last few years. I mod on Reddit and have seen this happen daily. They post a pro-nuclear/anti-renewables talking point and then when you look at the account history, the account will be days old or have a clearly artificial history generated by karma farming (reposting old comment or short junk comments generated by computer).

It follows much the same patterns we saw with accounts pushing climate change denial on social media. Research later showed this was driven by bots and paid trolls.

> Is there any reason to believe that this isn’t just an industry group with a friendly, activist face

I can almost guarantee this is the case. So-called environmentalist Micheal Shellenberger is another example (made more clear when he came out as a climate change minimizer recently). It uses the same playbook that fossil fuel companies used to create artificial climate change skepticism.

At least one group of reporters is doing investigative research into this issue.

One litmus test to use: most nuclear energy advocates that are acting in good faith see it as part of an all-of-the-above energy solution that includes nuclear, renewables, and storage. The shill accounts break from this by trying to trash renewables while pushing nuclear energy.

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
BS. Nobody who is involved with the energy sector thinks ElectricityMap has any credibility. The datasets are extremely fragmentary and have huge yawning gaps with no data available, which should be the first red flag for anybody citing it.

I have literally only seen ElectricityMap cited by nuclear energy shills -- as has been demonstrated in the comments here.

Edit: I won't deny it might have value for trends over time within a given country, but the country-to-country carbon emissions comparisons simply don't fit with official figures.

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
This is FALSE. The World Nuclear Industry Status Report[1] "estimates that since 2009 the average construction time for reactors worldwide was just under 10 years, well above the estimate given by industry body the World Nuclear Association (WNA) of between 5 and 8.5 years."

The most promising SMR project is already hitting the same delays and cost overruns that plague the nuclear industry. Citation: https://www.sciencemag.org/news/2020/11/several-us-utilities...

Quote: "announced that completion of the project would be delayed by 3 years to 2030. It also estimates the cost would climb from $4.2 billion to $6.1 billion."

1: https://www.reuters.com/article/us-energy-nuclearpower-idUSK...

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
The World Nuclear Industry Status Report[1] "estimates that since 2009 the average construction time for reactors worldwide was just under 10 years, well above the estimate given by industry body the World Nuclear Association (WNA) of between 5 and 8.5 years."

The data on ElectricityMap is dubious at best -- I'll give them the benefit of the doubt and assume it's simply due to not updating with more recent figures for some countries.

1: https://www.reuters.com/article/us-energy-nuclearpower-idUSK...

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
Worse, it's a total gamble on whether or not we'll even be able to BUILD a given reactor. Reactor construction projects have an absolutely gobsmacking rate of delays and massive cost overruns, and many end up getting canceled.

Flamanville in France is fully 3x its original cost estimate. If it manages to hit its current planned opening date in 2022, then that will be after 15 YEARS of construction.

Out of the recent wave of reactor construction in Europe and the US, most are have hit massive delays and cost overruns. At this point it's hard to conclude that the nuclear industry is remotely competent to deliver on any single promise.

CogentHedgehog··on Campaign for a Green Nuclear Deal: Forging a New America
The problem with nuclear energy is that it's 3-4x the price of solar/wind per unit energy[1], and it takes 10+ years to build most reactors.

If you want to cut carbon emissions ASAP it's important to pick the fastest, cheapest zero-carbon solution.

1: https://www.lazard.com/media/451445/grphx_lcoe-02-02.jpg

CogentHedgehog··on My Resignation from the Intercept
People that support political propaganda don't get a right to be offended by fact-checking in credible journalism. People that support fascist leaders don't get to complain about "censorship." Not now, not ever.
CogentHedgehog··on My Resignation from the Intercept
I phrased that poorly. Fascism uses economic and class features that largely do not exist in a far-left society, such as a sizeable middle class and corporate power.

From Umberto Eco's definition of fascism: "'Appeal to a Frustrated Middle Class', fearing economic pressure from the demands and aspirations of lower social groups."

From Gentile's definition: "corporative organization of the economy that suppresses trade union liberty, broadens the sphere of state intervention, and seeks to achieve, by principles of technocracy and solidarity, the collaboration of the 'productive sectors' under control of the regime, to achieve its goals of power, yet preserving private property and class divisions;"

I understand that it is appealing to try to dilute the meaning of "fascism" by applying it overly broadly until it loses its meaing. Whataboutism is also convenient way for the Right to distance themselves from the evils committed by the right-wing Nazis and other far-right fascist groups.

The better way for the Right to distance itself is by supporting moderate democratic norms rather than giving in to violent populist impulses. Power is one heck of a drug though.

CogentHedgehog··on My Resignation from the Intercept
A lot of the commentators don't WANT sane publications. For this exact reason. They like that candidate and would rather burn sane journalism down then see him lose.
CogentHedgehog··on My Resignation from the Intercept
My conclusion is making sense of both of these.

Why didn't I believe Greenwald initially? Well, I have had the dubious pleasure of seeing many people behave badly and then play the morally-outraged victim to try to legitimize it. There's a particular way they write and talk, and Greenwald's description of events reeked of it.

The Intercept's response made a few jabs but it didn't level any particular wild accusations, unlike Greenwald. It seemed much more credible, although clearly frustrated. I assume (reading between the lines) there he developed a prior history of conflicts and bad blood at the Intercept, and political differences. This kind of blow-up does not just come from nowhere.

Greenwald has since released the emails, and they sound much closer to what the Intercept said than what he did: https://greenwald.substack.com/p/emails-with-intercept-edito...

The editors clearly had substantive journalistic concerns about the extent Greenwald was drawing conclusions from dubious evidence, and the lack of context around it. They're comfortable publishing the parts of the story he can substantiate, but not the parts that are weaker (or less on-topic). There have some political disagreements, but that's hardly censorship.

CogentHedgehog··on My Resignation from the Intercept
You clearly know how easy it is to take someone's words out of context and twist them into something they aren't. A good editor brings clarity and brevity.

The editor was telling Greenwald to focus on the story he could prove and tell effectively. That's not "censorship", that's good writing advice. He wanted to propagate flimsy partisan conspiracies instead.

If he wants to do that in his own name, that's fine. But he is not entitled to put the Intercept's reputation behind it.

CogentHedgehog··on My Resignation from the Intercept
Sure, there's TONS of things that are 'not fascistic' but are bad.
CogentHedgehog··on My Resignation from the Intercept
That's not true either though. Fascism is by definition a right-wing form of authoritarianism, because of how it uses class divides to build its political power.

The corresponding left-wing authoritarianism would be "leftist populism", "Stalinism", or some other term.

What you're describing is called Horseshoe Theory, and it's not well-supported by evidence. Right-wing and left-wing populism have a few things in common (authoritarianism primarily), but have much larger differences than similarities.

I will point out that attacks on a free press ARE a clear sign of authoritarian regimes. I would encourage you to take a look at the history of the German word "Lügenpresse" (literally "lying press"). Especially take note if you see similar things being said about the modern media -- and which groups are saying it.

CogentHedgehog··on My Resignation from the Intercept
I will agree that people intentionally misuse the word "fascism."

For example, the people in this comment thread claiming that editors making editorial decisions are the "fascist left" are pretty clearly misusing the word.

CogentHedgehog··on My Resignation from the Intercept
If anything, Greenwald comes off as compromised by his own political biases. The editor says fairly gently that there are concerns with poorly-supported claims and omitting key information -- such as the media reporting on corruption claims and the fact that media outlets have not been given access to the supposed hard drive of emails. This results in a meandering and potentially misleading article that leads readers to conclusions it does not actually support.

They suggest a way to focus it more on core points which are well-substantiated and get a solid article out of it. They're not censoring his political views, if anything they're encouraging him to express them in an article indicting liberal media for going soft on Biden. They're trying to get a shorter, more tightly-focused article for publication. Which is to say, a more solid article.

That sounds like an editor doing their job -- editing is supposed to be the art of removing words, after all!

Greenwald sends first a polite reply, then a much less polite one that jumps to this vitriolic claim:

> I want to note clearly, because I think it's so important for obvious reasons, that this is the first time in fifteen years of my writing about politics that I've been censored -- i.e., told by others that I can't publish what I believe or think

Followed by insinuations that they're suppressing the story due to their political biases. It sounds like Greenwald can't accept that there may be legitimate explanations for why the content isn't focused or solid enough.

That's downright nasty and unprofessional.

CogentHedgehog··on My Resignation from the Intercept
I'd bet on bot-busting or possibly something that de-ranks stuff with political keywords close to the election to avoid spam and noise. Human moderation and content curation is a lot rarer than people think online. Primarily because it takes a lot of labor and and response times tend to be slow compared to how fast things go viral.
← PreviousPage 2 of 4Next →