Lucy
5 followers · 516 views
on the atlas · 38
- Writing an OS in Rust5 savers
- A Survival Guide to a PhD28 savers
- Applying to Grad School: Programming Languages Edition3 savers
- To academia or not to academia7 savers
- dalle-2-preview/system-card.md at main · openai/dalle-2-preview2 savers
- What even is compiler correctness?1 savers
- Self-reported long COVID after infection with the Omicron variant in the UK - Office for National Statistics1 savers
- Turing is Useless1 savers
- What Every C Programmer Should Know About Undefined Behavior #3/3 - The LLVM Project Blog1 savers
- What Every C Programmer Should Know About Undefined Behavior #2/3 - The LLVM Project Blog1 savers
- What Every C Programmer Should Know About Undefined Behavior #1/3 - The LLVM Project Blog1 savers
- Developing Developers1 savers
- History of Programming Languages1 savers
- The Design Recipe1 savers
- 30 Best Math Books to Learn Advanced Mathematics for Self-Learners1 savers
- How Not to Teach Recursion2 savers
- Welcome! | plai1 savers
- Brown CS: CSCI 1730: Programming Languages1 savers
- What is a Pedagogic IDE?1 savers
- FatNums: An Alternate Representation for School Arithmetic1 savers
- Scribble: The Racket Documentation Tool1 savers
- The Role of Computer Science in Computer Science Education - YouTube1 savers
- Shriram Krishnamurthi: Talks1 savers
- Molecular Computing1 savers
- A battle is raging over long covid in children1 savers
- SARS-CoV-2 aerosol risk models for the Airplane Seating Assignment Problem - ScienceDirect1 savers
- This startup wants to make electronics out of single molecules1 savers
- Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing1 savers
- The Evolution and Biology of SARS-CoV-2 Variants1 savers
- Ali Alkhatib: Digital Forests1 savers
- Computers can be understood14 savers
- A Conversation with David Shaw - ACM Queue1 savers
- A Conversation with Margo Seltzer and Mike Olson - ACM Queue1 savers
- Medical AI Safety: We have a problem. – Lauren Oakden-Rayner1 savers
- Medical AI Safety: Doing it wrong. – Lauren Oakden-Rayner1 savers
- Fools and their time metaphors9 savers
- On the Phenomenon of Bullshit Jobs4 savers
- Alan Kay4 savers
highlights · 304
is of biosensors that would enable people to check in on biomarkers like vitamin levels or evidence of an infection with only a little more hassle than it now takes to check their heart rate on a smartwatch.
This startup wants to make electronics out of single molecules“They can encode all these interesting different states, like in biochemistry, that we just don’t have other ways of accessing.”
This startup wants to make electronics out of single moleculesMolecules are special because “they can be so much more complex than binary,”
This startup wants to make electronics out of single molecules“self-assemble” circuits, adding molecules under highly controlled conditions and letting them assemble into the desired structures all by themselves
This startup wants to make electronics out of single moleculesAt that time, it seemed like the small size of molecules could help make circuit components tinier and computational chips denser
This startup wants to make electronics out of single moleculesits electrical conductivity changes
This startup wants to make electronics out of single moleculesmolecule to the other end of the wire
This startup wants to make electronics out of single moleculesmolecular wire made from a chain of amino acids that is connected to the rest of the chip just as a regular metallic wire would be
This startup wants to make electronics out of single moleculesIf we were doing drug development, we have only done the equivalent of simulation studies or animal models.
Medical AI Safety: Doing it wrong. – Lauren Oakden-RaynerSo far, no-one has ever shown that patients are better off when we use an AI system
Medical AI Safety: Doing it wrong. – Lauren Oakden-Raynerwe are not very careful with our study designs
Medical AI Safety: Doing it wrong. – Lauren Oakden-Raynerwe assume good experimental performance equals good clinical performance
Medical AI Safety: Doing it wrong. – Lauren Oakden-RaynerThey are treating these systems as simple medical devices, allowing decision-making AI systems to be used clinically with only modest evidence backing them and no investigation of the possible unintended consequences
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerBut we have to recognise that even the most cautious people can make decisions that lead to tragedy, which is why we have regulation in the first place
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerMany promising treatments have been abandoned because they caused more harm than they prevented
Medical AI Safety: We have a problem. – Lauren Oakden-Rayneroff-shoring to avoid testing altogether
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerFinally, several companies (i.e., Google, Enlitic, among others) have been practising “off-shoring” – they are taking their AI products to jurisdictions with more lax regulatory environments.
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerThis doesn’t mean we ignore these risks, just that the risk of disaster is fairly low.
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerCAD could lull the novice reader into a false sense of security. Thus, CAD had both lower sensitivity and lower specificity, a non-redeeming quality for an imaging test.
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerWhile it is not impossible that application of these systems will harm patients, it is unlikely.
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerThese systems all had something in common: a human expert is still required to make every decision.
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerAI applied to human medicine is different than AI in almost all other areas
Medical AI Safety: We have a problem. – Lauren Oakden-RaynerBut are we sure Facebook, Twitter, YouTube, Google, and all of these other companies aren’t, in their own ways, cultivating digital monoculture forests that are barreling toward disaster in the process of making the world digital?
Ali Alkhatib: Digital ForestsBut it was a whole other thing when they turned that simplifying function back onto the world, now shaping the world according to the reductive functions they used to interpret it. When you reshape the world to reflect the square you drew, you invade and disrupt all of the lives that interact with that forest - not just the birds and insects that make homes in those trees, but the people that live nearby and hunt (or live, or just hang out) in that area as well.
Ali Alkhatib: Digital ForestsI also have a remarkable capacity to delude myself into thinking that each idea has a higher probability of working than it really does, which provides me with the motivation I need to keep working on it.
A Conversation with David Shaw - ACM QueueI tend to generate a plentiful supply of ideas, the vast majority of which turn out to be bad ones.
A Conversation with David Shaw - ACM QueueSince my own interest is in the application of molecular dynamics simulations to biological problems, I haven't been forced to think very hard about what aspects of the approach we've followed might be applicable to the co-design of special-purpose machines and algorithms for other applications.
A Conversation with David Shaw - ACM QueueWe were also able to make use of problem-specific knowledge to provide hardware support for a specific type of inter-chip communication that was especially important in our application.
A Conversation with David Shaw - ACM QueueSome people think all advances in computing are due to software or hardware, but I think some of the most interesting ones are where those things co-evolve in some sense.
A Conversation with David Shaw - ACM QueueForce calculations between particles are such an important part of computational science, and to have made the observation that the best way to compute the interactions of two particles involves sending them both somewhere else is just amazing.
A Conversation with David Shaw - ACM QueueThe one I'm referring to is the NT algorithm, which is what runs on Anton. The midpoint algorithm is used in Desmond, an MD code that was designed to run on ordinary commodity clusters.
A Conversation with David Shaw - ACM Queuesimulate the motions of large biomolecules, such as proteins or DNA, at an atomic level of detail over a simulated time period during which biologically interesting things happen in real life
A Conversation with David Shaw - ACM Queuebut the applications that really drive me are biological, especially things that might lead not only to fundamental insights into molecular biological processes, but also to tools that someone might use at some point to develop lifesaving drugs more effectively.
A Conversation with David Shaw - ACM Queuebringing a bit of computer science into a research area where there hadn't yet been all that much of it.
A Conversation with David Shaw - ACM Queue"Oh, that's really just a special case of this known problem, with the following wrinkle."
A Conversation with David Shaw - ACM QueueRich is a computational chemist, so a lot of what he did involved algorithms, but his training was in chemistry rather than computer science, so when we got together socially, we often talked about some of the intersections between our fields.
A Conversation with David Shaw - ACM QueueRich Friesner, who was a chemistry professor at Columbia
A Conversation with David Shaw - ACM QueueAnton, a special-purpose supercomputer designed to speed up molecular dynamics simulations by several orders of magnitude.
A Conversation with David Shaw - ACM QueueWe try to take out or refactor a significant amount of code to keep the overall size down.
A Conversation with Margo Seltzer and Mike Olson - ACM QueueWe were opinionated nerds who had driven strategy on the basis of our engineering taste. We wouldn't be in a position to do that in the same way with a traditional venture capitalist calling the shots in the boardroom.
A Conversation with Margo Seltzer and Mike Olson - ACM QueueIt also created a very different mentality. Since every person in the company stood to gain by every dollar of profit the company made, the employees treated company money like their own money. You didn't see our sales team spending ridiculous amounts of money or making promises for features that engineering couldn't deliver.
A Conversation with Margo Seltzer and Mike Olson - ACM Queue"Things that happen once in a million times happen millions of times every day for us.
A Conversation with Margo Seltzer and Mike Olson - ACM QueueWe had a really great, reliable, single-node system. But if your single node crashed, then your service was down. By the early 2000s, that was no longer acceptable.
A Conversation with Margo Seltzer and Mike Olson - ACM QueueThe demand for the embedded market was really driven by these "Web 1.0" companies. That was where Sleepycat really made it. We became the back end for pretty much every first-generation web service.
A Conversation with Margo Seltzer and Mike Olson - ACM QueueBecause of the Netscape money that started the business, we never got addicted to venture capital. We were running lean. We were profitable from the very beginning.
A Conversation with Margo Seltzer and Mike Olson - ACM Queue“Ultimately you're going to have the state colluding with these private companies, because the state has no capacity to run such a massive complicated network of CCTV cameras,”
South Africa’s private surveillance machine is fueling a digital apartheidAs a result, it’s predominantly white people who have the means to pay for surveillance, and predominantly Black people who end up without a say about being surveilled
South Africa’s private surveillance machine is fueling a digital apartheidBut absent from the conversation is why the crime exists in the first place.
South Africa’s private surveillance machine is fueling a digital apartheidTaiwan and South Korea, people can go see a doctor without worrying about crushing medical debt. Paid sick leave is a guarantee. People use masks and when the government tells them what to do, they generally trust their political leadership not to be corrupt or incompetent.
Ali Alkhatib: We Need to Talk About Digital Contact TracingThe second is that the way the US has cleaved a boundary between wealthy and poor
Ali Alkhatib: We Need to Talk About Digital Contact Tracing