Hacker News Reader

ampdepolymeraseon Clumps of human brain cells in a dish can learn to play Pong faster than an AI

This is not the first attempt. Growing tiny brain organoids is fairly well understood, scaling it up is not. I have not read the paper but previous attempts at this have struggled with getting useful outputs out of the neural culture, most of the time the spikes just add up to noise.

ampdepolymeraseon Abolish Yale

Same quality yes, but not access to the same opportunities.

ampdepolymeraseon Abolish Yale

Tertiary education at the elite level is rarely about the quality of education. The difference between Georgia Tech and Columbia or Harvard's CS program is not that stark. However, the story is completely different when it comes to alumni network, access to funding, future opportunities. Prestige and network effects compound and an argument can be made for whether we should allow academia to be the gatekeepers of social mobility at the highest level.

Certain Ivies favour grade inflation yet their students do not suffer dilution of their degree value, this is not a privilege extended to many other institutions.

ampdepolymeraseon NewLimit: a company built to extend human healthspan

> What differentiates SpaceX from NASA, or SpaceX from Blue Origin, is people and culture. We are not trying to build an institute or academic minded organization where papers are more important than products. Our goal is to build an ambitious, well run, for-profit company that will deliver revenue generating products on the way toward accomplishing its much bigger objective.

It is debatable whether they truely understand the financials of biotech. The grind of basic research will never go away. Many successful biotech company essentially acquihire researchers and work that is already 80% complete, the role of the company is to bring it to production. That itself will make up the bulk of the company's workload. Similarly, SpaceX had the benefit of leveraging an existing pool of talent and resources; you cannot build a heavy launch company in Zimbabwe. If they want to do both active basic research and at the same time trial therapies, then they would need an enormous amount of funding (on the software unicorn level). You will need scale on the same magnitude as Pharma giants like Pfizer, GlaxoSmithKline, Johnson and Johnson to be able to acquire companies, run trials, and discard ideas that do not work. In the current bull market, this is the perfect company to build with Coinbase's founder as the chief fundraiser.

ampdepolymeraseon Tiny primate may join the ranks of the world’s model organisms

We usually use macaques for this (I haven't read the paper). Is there a review paper comparing both creatures? If they are really better, I should look into procuring some. Macaques are quite expensive in north America.

ampdepolymeraseon Tiny primate may join the ranks of the world’s model organisms

Imagine neutering your dog/cat so they would be a less aggressive pet.

ampdepolymeraseon DangerousThings Biohacking Forum

They ought to rename themselves to biomedical hacking. Biohacking in contemporary usage usually refers to homebrew laboratory-grade biotechnology (those involving PCRs, gene modification, a ton of biochemistry). The dangerous things forum is primarily about implants which falls under biomedical and neural engineering, not biohacking.

On a side note, I strongly do not recommend trying out any of the biomedical implants without a good reason. You can approximate magnet implants to a reasonable extent with a bit of superglue and RFID chips can be placed in rings and jewellery. Risk of infection aside, accidental nerve damage is very difficult to fix with our current level of understanding.

ampdepolymeraseon Lead contamination in hunted meat

Be careful of prions, wild game is not the safest source of protein.

ampdepolymeraseon RNA Takes Over

Molecular biology education needs more funding. Teaching it at scale (and more importantly, communicating new discoveries outside of textbooks) still remains a challenge. Keeping up with the field without reading papers is much harder compared to CS/ML where almost every ML engineer maintains a blog.

ampdepolymeraseon Crohn’s Disease: Is the Cold Chain Hypothesis Still Hot?

If tube fed diets are the trick, why not just go with Soylent? It's about as bacteria-free as food can reasonably get.

ampdepolymeraseon Mushroom leather and how it is made (2020)

I did say with the caveat of chemical treatment.

ampdepolymeraseon Mushroom leather and how it is made (2020)

Anything biodegradable (in the composting sense) almost always does not last long absent significant chemical treatment. If you think current faux-leather technology is a step down from the real thing, then mycelium based products will be even worse. There is an inherent trade-off between biodegradable and long lasting products. What we can do is make products that are recyclable. In other words, engineer materials to only break down with a specific brand of artificially engineered enzymes that are not found in the wild. Unfortunately we are still a couple years away in terms of proteomics advancement to be able to do de novo synthesis with such precision.

ampdepolymeraseon Team builds first living robots–that can reproduce

At about the same time as when growing organs in test tubes becomes viable and restaurant grade synthetic wagyu steaks grown from stem cells become reality.

ampdepolymeraseon Team builds first living robots–that can reproduce

I have commented on this before so I will link it here:

https://news.ycombinator.com/item?id=27332700

For xenobots to be useful, the underlying cell behaviour needs to be better understood and controlled. Their research appears interesting if you are a computer scientist or physicist. But for practical applications and applied biotechnology what they are doing is an curious gimmick at best. If they can perfect their understanding, then they would have basically solved half the puzzle of organ engineering. But right now it is no different from black box engineering.

This is very similar to Turing's reaction-diffusion models on leopard spots — great for computer modeling and popular science magazines, not so immediately useful if your grandmother gets cancer.

I will admit it looks perfect to sell as a biotech startup to dumb SV VCs who don't do their due diligence when it comes to investing in applied science fields.

ampdepolymeraseon Ask HN: What is something important I should start doing in my 20s?

You cannot change the latter without having some effect in the former because of basic mathematics of ratios.

ampdepolymeraseon Ask HN: What is something important I should start doing in my 20s?

Crunches builds abdominal muscle but body weight is affected only by cardio.

ampdepolymeraseon Fingerprints can be hacked

OP was referring to state level actors.

ampdepolymeraseon Fingerprints can be hacked

Why in the world would you need CRISPR or lab grown meat? Just sequence the DNA and send it off to a DNA assembly service. The price is a couple hundred bucks a pop. You don't have to replicate the entire DNA, just the segments used for forensic PCR.

(On a side note, the state of biotechnology and life science knowledge on HN is utterly deplorable, repeating buzz words does not reality make.)

ampdepolymeraseon Solar-driven water splitting at 13.8% solar-to-hydrogen efficiency

Elemental hydrogen has this problem: it is hard to store. You need low temperatures, high pressure, and corrosion resistant containers. But there is one weird trick: if you bind hydrogen to carbon atoms, then you get best of both worlds. The fuel remains energetic and yet is relatively unreactive outside of combustion and is easy to store. Perfect for aviation.

ampdepolymeraseon Skeptical about electric-powered planes

If you want to take a plane without constant heating, be my guest.

ampdepolymeraseon RF Magnetoelectric Antenna for Neural Activity Monitoring

Most commenters here are misunderstanding the paper. This is research on a wireless RF power transfer and low strength magnetic field sensing system for BCI sensors. It is not about BCI sensors itself. This is about wireless power transfer and modulating the reflected signal to broadcast the sensor information. In other words your brain computer interface no longer needs a subdermal battery.

ampdepolymeraseon All Those 23andMe Spit Tests Were Part of a Bigger Plan

One was a university student, the other had immediate access to Google's founders at the highest level. This isn't a fair comparison.

ampdepolymeraseon Isomorphic Labs: Reimagining drug discovery with an AI-first approach

Protein catalysts almost always contain an inorganic atom. We will be able to make more efficient catalysts with better understanding of protein structure but the need for metals won't necessarily go away.

ampdepolymeraseon C³: A "cool" route to the Higgs boson and beyond

It has created a ton of jobs for instrumentation engineers and PhDs. It has also brought many economic benefits to Switzerland.

ampdepolymeraseon Intel Extension for Scikit-Learn

The editor autocompletion and standard library documentation could use a lot of work. The introductory tutorials are overly focused on type theory and details and do not give a good overview of which generic data structures to use in production code. Julia's JIT is very different from other conventional mainstream languages and the process of selecting standard library generic data structures for optimal performance is very poorly documented.

There is no Effective Julia style of guide. You either have to wade through infantile tutorials for those with minimal programming experience or several reference books worth of nitpicking on syntax. The actual methods themselves are not well documented and lack examples and usage guidelines.

The language and ecosystem do not feel like a project backed by commercial funding, it feels like one of those functional languages out of academia research where the structure and design of the language are more important than actual developer experience. There are many new projects but most are not actively maintained and updated. The language itself feels massive, with syntactic sugar and weird types everywhere. Trying to understand the implementations of other people's Julia code is frustrating, similar to reading a library written in pure C++ templates. Compared to Go/Rust/Dart, Julia feels overly convoluted. Julia literature is structured in a way that seems to heavily encourage you to take regular classes and lectures to learn and pick up the language. It is hard to feel productive from the get-go.

ampdepolymeraseon Animated guide to Symex: Emacs structural editing with Lisp

Is this designed to work together with Evil mode? The keybindings look to be Vim based.

ampdepolymeraseon Ask HN: What are the problems that are willing to pay more than $1000 to solve?

And they are the reason why we have a housing and healthcare crisis.

ampdepolymeraseon Subsidizing costs while restricting supply drives America’s fiscal imbalance

The residency system needs to go. This is the major bottleneck. Too much innovation is held back because the supply of doctors is just too constrained via education through apprenticeship. If the only profitable job in life science is medicine (the physician track), then all those equally talented people who lose the rat race for one reason or other would be forced to seek other careers.

ampdepolymeraseon DIY Acoustic Camera

There is a long trail of dead startups attempting this. But don't let this dissuade you, please do a Show HN when you launch.

ampdepolymeraseon Plans to shift aviation from magnetic navigation to true navigation

You need clear view of the sky and big lenses (you can always use more compact metamaterial lenses and infrared imaging for cloudy days but they are as expensive if not more than INS). Star charts also drift over the years and you will have the same problem as magnetic true north where the database needs to be updated regularly. Fine for airliners, not so great for skyhawks.