maxander

Pretty few would take the risk, I bet. It's an interesting problem- we have "trigger warnings" and "content warnings" but not "content all-clears."

It's definitely a real thing. I don't have it badly, but when I learned about tryptophobia it was satisfying to know why lotus seed pods had always made me feel somewhat uncomfortable.

(Google that only if you're reasonably confident you're not a tryptophobe. :) )

Trypophobia is presumably an abnormal, overly-aggressive version of the disease-avoidance mechanism, though; there's no evolutionary reason to be unsettled by holes on inorganic objects, or even most kinds of holes on organic ones. Presumably, almost everyone has some level of the underlying mechanism; go look at a picture of a smallpox victim and see how you feel about it.

This sort of story probably fully explains origins of the OP's lament (that AWS billing is so opaque that even internal Amazon teams can't figure it out.) Some AWS middle manager knows that if he orders better billing tools to be implemented, AWS income will drop; not significantly, but ~$700 multiplied by the number of AWS users might make enough of a dent to be noted in his performance report. In order to make this omission more plausible, even internal investigation tools (which would encourage infrastructural improvements that make user-facing tools easier to build) are neglected.

If there are things we can't conceptually comprehend, then definitionally we don't understand what they are, so of course we can't answer this question. We won't know one way or another unless we consult with something a step or more "above us on the scale" and they tell us.

(And even then- how do we judge if something is "above" us? Chimps would have about as much trouble understanding IQ or the even the abstract notion of "intelligence" as they do skyscrapers, I'd imagine.)

Academia, for all its faults, is still an excellent system for collecting those who are obsessive about s subject (at least for a certain set of “established” subjects.) It’s preposterously difficult to do the PhD-through-tenure course if you’re not obsessed, and it certainly isn’t the financially optimal career path; if someone has made it through that system, you can be sure they’re happy sitting and doing their subject all day, for days.

In about 10 years, it will be possible to run a cell lysate sample through a mass spec machine and get what a present-day scientist would call a pretty good picture of everything going on at the proteomic/metabolomic level, in perhaps ~5 hours.

But in that time we will have probably discovered several currently-unappreciated, biologically relevant biochemical mechanisms which we can’t efficiently probe like this. And also it will be considered next to useless because it doesn’t work on single-cell samples. :)

> “I would hate to see say a low-income family who’s concerned about air [quality] review their options and say ‘I could either buy a $400 air cleaner, or I could go out and buy a $30 plant,’” says Gall. “That plant is not going to improve their air quality—full stop. It just won’t.”

The real question is, why does the air cleaner, which is usually just a fan blowing through an activated charcoal filter, cost $400 in the first place? There’s a market to disrupt, surely.

A dedicated, well-maintained name abstraction is certainly something that needs to happen. More than interesting, it's a bit bizarre that this hasn't been done yet (AFAIK.)

In terms of developer-facing complexity, this could be a laughably simple thing to use- just a type that supports equality, perhaps ordering, and conversion to string. Only the constructor would need to be complex. :)

I guess the reason this hasn't been done is simply that the implementation would never be "correct"- there is no formal specification of human names out there, and there would always be cases where some poor individual with an unusual name falls afoul of the system. Strictly fewer cases than we have today, where everyone rolls their own name system, but still some; it's not a solvable problem the way timezone conversion is.

But, on the other hand, that's the way the world is- messy. Developers are going to have to learn better best practices for communally doing our best in cases where there is no perfect answer, because there's only going to be more of such cases as tech continues to eat the world.

Gardasil is described as a prophylactic vaccine, rather than therapeutic, so no- it sounds like you would have to be vaccinated prior to infection in order for it to be effective. However, it sounds like there are therapeutic vaccines in development [0, 1].

[0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3586871/ [1] https://www.ncbi.nlm.nih.gov/pubmed/30955915

SEEKING WORK

  Location: Boston, MA
  Technologies: Python, C/C++, C#, Linux 
  Résumé/CV: https://drive.google.com/file/d/1vCk9BIAJHF1hQnV071JxBZ6dGTR2Hb7G/view?usp=sharing
  Email: w.max.alexander@gmail.com

Six years of experience working as a bioinformatics engineer and looking to explore new areas in the biotech space. My educational background includes enough biology that I can read a paper and understand an experimental setup, but my skillset is a strong mix of software engineering and data science; I can build the two-way interface between your lab and the data center.

  Location: Boston, MA
  Remote: Possible
  Willing to relocate: Possible
  Technologies: Python, C/C++, C#, Linux 
  Résumé/CV: https://drive.google.com/file/d/1vCk9BIAJHF1hQnV071JxBZ6dGTR2Hb7G/view?usp=sharing
  Email: w.max.alexander@gmail.com

Six years of experience working as a bioinformatics engineer and looking to go further in the biotech space. My educational background includes enough biology that I can read a paper and understand an experimental setup, but my skillset is a strong mix of software engineering and data science; I can build the two-way interface between your lab and the data center. Looking for full-time or contract employment.

Which means the root password is probably still “admin” :)

I'd argue the opposite- it's like saying that water is legal but it's illegal to drown someone. Nothing in the field of machine learning is in itself illegal, as opposed to things like hard drugs and heavy weaponry, which are in and of themselves illegal for private citizens, regardless of what they're doing with them.

While we're at it, whenever anyone talks about machine learning they're not talking about a technique of mathematics, they're talking about a technique in software development (albeit that has some mathematical roots.) It's not popular because we're all passionate about discovering mathematical truths, it's popular because we can use it to make computers do cool things. :)

There’s still no regulation on what you can do with machine learning (At least until your deepfakes constitute unlawful use of someone’s likeness, or you use your classifier to make hiring decisions or something, or something like that.)

Location: Boston, MA

Remote: Yes

Willing to relocate: No

Technologies: Python, Haskell, R, C/C++/C#; machine learning and statistics.

Resume/CV: https://www.linkedin.com/in/william-max-alexander-8940469/

Email: w.max.alexander@gmail.com

Master's degree in computer science and 5 years of experience in bioinformatics. My bioinformatics experience centers on data processing and analysis workflows for proteomic mass spectrometry experiments. I'm also interested in working with deep learning and other advanced machine learning techniques for biology/biotech applications.

Open to either full-time employment or consulting opportunities.

Bias in AI (whatever the source or nature of the problem) is a real issue that needs addressing, but taking the relevant training data off of ImageNet seems like a perfect example of papering over a problem to avoid really confronting it. We will need to find ways to make AI programs than can see beyond the biases (about humans or otherwise) that will always exist to some degree in real-world data.

If the ImageNet contains bias that leads to embarrassing results- that’s fine! That give us a readily available toy instance of the problem to study. Taking that away could actively harm anti-bias research.

If you you want “affordable rents,” you have two options; high density or long commutes. And high density means large buildings. There’s just geometrically no way to supply a SF/NYC/Boston level city primarily through detached homes.

It’s on cities to solve the problems you describe, usually through housing inspection and tenant rights law.

I’ve never been to Russia and don’t know the Russian attitude towards tech or voting. But I’m confident that if the US had a system like this and made a reasonably usable phone app to be the front end, almost no one under the age of 60 would bother to wait in line for a physical voting booth ever again.

As far as I understand it, the mechanisms that trees and other plants use to make themselves plant-shapes wouldn’t generalize to other useful things like houses; e.g., the algorithm is something akin to an L-system [0], and there’s only so much one can do with that.

[0] https://en.m.wikipedia.org/wiki/L-system

But most of these students (particularly ones taking the lin alg course that apparently has APPLIED in all caps in the title) are going to be engineers or similar who will never encounter the fundamentals outside of their application in tools like k means. Teaching them an abstract field without connections to what motivates it (intellectually or practically) is pedagogically not ideal.

I wonder if the issue is liability- people suffering from various ailments are liable to Google their symptoms in hopes of finding an at-home cure instead of going to a doctor[0], and if this turns out badly they could conceivably turn around and try to sue Google for providing harmful medical advice[1]. De-ranking sites that provide medical advice, which aren't whitelisted well-known sites like Mayo Clinic or WebMD, might prevent this- or at least provide a better defensive position in the courtroom.

[0] Especially likely in countries with badly-structured healthcare systems, such as the U.S.

[1] Especially likely in countries with badly-structured legal incentives, such as the U.S.

I got just the opposite impression- the interviewer gave a problem that is different from the normal “sorting problem,” particularly so that the answer wouldn’t be any of the algorithms one learned in school. He was presumably hoping that the interviewee would get that, figure that the added constraints allowed a more efficient solution, and try to think through a novel algorithm. If the guy had used a bit of creativity, it wouldn’t matter if he didn’t know radix sort from a hole in the ground.

I couldn’t promise that I’d make that leap either, sitting sweaty-palmed in front of a hiring manager, but I see what the interviewer was thinking.

It's for the better- even ignoring the risk of physical harm, these motors are loud. Cities would be orders of magnitude more obnoxious if you had the upper class constantly buzzing overhead in vehicles that made a Harley-Davidson sound like a soft purr.

Two answers:

- Hygeine? That was invented by Louis Pasteur in the nineteenth century. :)

- In circumstances that allowed it, bread was often used as a plate (“trenchers.”) Would be nicely hygienic if the bread used in this way wasn’t traditionally then given as alms.

Location: Boston, MA

Remote: Yes

Willing to relocate: No

Technologies: Python, Haskell, R, C/C++/C#; machine learning and general statistics

Resume/CV: https://www.linkedin.com/in/william-max-alexander-8940469/

Email: w.max.alexander@gmail.com

Master's degree in computer science and 5 years of experience in bioinformatics. My bioinformatics experience centers on data processing and analysis workflows for proteomic mass spectrometry experiments, but I'm interested in other "omic" technologies. I'm also particularly interested in working with deep learning and other advanced machine learning techniques for biology/biotech applications.

Open to either full-time employment (starting September or October '19) or consulting opportunities.

All sorts of medical drugs have contraindications- that’s why we have doctors prescribe them in the first place. E.g., you have severe sleep apnea, you shouldn’t use benzodiazepines; if you have bad anxiety you shouldn’t take address; etc etc. LSD will take a bit more diligence (it’s powerful!) but is not qualitatively different in that respect.

Natural human grammar is error correcting codes, essentially. If sentence should e corrupted, e.g. like this one is, it usually is immediatly visible that is wrong, and what the “correct” should be.

IIRC, one of the criticisms of Ithkuil is that there’s so much semantics packed into the syntax that this doesn’t work for it- a corrupted Ithkuil sentence often turns out to be a valid sentence with a different meaning.

Nee gig for Musk’s Boring Company? Drilling water pipelines right through the mountains.

No good reason- but like I said, check out the price tag; a mile of subway development in the U.S. can cost billions (with a B!) of dollars. Additional foot traffic along that mile couldn't get you that much back.