You can emulate the Victor 9000 / Act Sirius 1 computer he talks about in the article with mame. There's a lot of software on the internet archive or this machine. It's quite interesting to look at the hardware level how many differences there are in the design decisions compared to an IBM PC.
Go to a modern hospital emergency room, it's a cacophony of devices all vying for attention. I walked down the hallway and realized every room in the place had a different audible alarm—all active! I suspected the device manufacturers were all worried about liability for their device, making sure to notify that a patient had a problem. The end result for the medical staff was an endless chaos of noise. Complete systemic failure of UX from a practical standpoint.
Yes. I have a family member that has had many hospital stays over the last few years, and one of the most obnoxious things is that the staff just lets everything beep. The last time we were in the emergency room the blood pressure monitor did not work and the staff didn't notice for over an hour. Even when it does work, they're constantly in an alarm state because patient has chronic high blood pressure. They either can't or won't silence the alarms, so every room is beeping, the nurse's station is beeping, their phones are beeping, and it's all being ignored. It's the very definition of alert fatigue.
In the regional hospital near me, they've begun actively fighting for fewer alarms. In part because they annoy everyone: patients, visitors, and hospital staff alike. But mostly because the inevitable alarm fatigue that the cacophony results in actively endangers patient safety.
The policy of this hospital is that all alarms, beeping, etc. should be disabled except in limited circumstances. Particularly at night.
From time in hospitals I've gotten very good at disabling them. Most nurses are fine with it but every now and then one would come on shift and tut tut at me for having done it. They usually shut up when I point out that they don't respond to the alarms in any sort of prompt way - as I'm sure if I were to continue pointing that out up their chain of command they would then find some trouble.
I always tell people though that being in the hospital doesn't make you healthier, mainly because you can't sleep. The hospital should be the absolute last resort, and your first priority on finding yourself in one should be to figure out how to get out of it, even if it involves nursing care at home.
And in my experience (not surprisingly) they have all developed a good sense of what alarms can be ignored, so like a pump beeping because it's done delivering some medicine doesn't matter so they ignore it and let it beep, but it matters to the parents with new baby trying to get some sleep.
The poster's looking for articles, so this recommendation's a bit off the mark. I learned more from participating in a few Kaggle competitions (https://www.kaggle.com/competitions) than I did from reading about AI. Many folks in the community shared their homework, and by learning how to follow their explanations I developed a much more intuitive understanding of the technology. The first competition had a steep learning curve. I felt it was worth it. The application of having a specific goal and the provided datasets made the problem space more tractable.
Not the poster you responded to but I learned quite a bit from kaggle too.
I started from scratch, spent 2-4 hrs per day for 6 months & won a silver in a kaggle NLP competition. Now I use some of it now but not all of it. More than that, I'm quite comfortable with models, understand the costs/benefits/implications etc. I started with Andrew Ng's intro courses, did a bit of fastai, did Karpathy's Zero to Hero fully, all of Kaggle's courses & a few other such things. Kagglers share excellent notebooks and I found them v helpful. Overall I highly recommend this route of learning.
I started with this 3 part course - https://www.coursera.org/specializations/machine-learning-in.... I think the same course is available at deeplearning.ai as well, I'm not sure, but I found coursera's format of ~5 min videos on the phone app very helpful (with speed-up options). I was a new mother and didn't have continuous hours of time back then. I could watch these videos while brushing, etc. It helped me to not quit. After a point I was hooked & baby also grew up a bit and I gradually acquired more time and energy for learning ML. :)
fastai is also amazing, but it's made of 1.5 hour videos, and is more freeflowing. By the time I even figured out where we stopped last time, my time would sometimes be up. It was very discouraging because of this. But later, once I got a little more time & some basic understanding from Andrew Ng, I was able to attempt fastai.
i mean yes but also how much does kaggling/traditional ML path actually prepare you for the age of closed model labs and LLM APIs?
im not even convinced kaggling helps you interview at an openai/anthropic (its not a negative, sure, but idk if itd be what theyd look for for a research scientist role)
I learned ML only to satisfy my curiosity, so I don't know if it's useful for interviewing. :)
Now when I read a paper on something unrelated to AI (idk, say progesterone supplements), and they mention a random forest, I know what they're talking about. I understand regression, PCA, clustering, etc. When I trained a few transformer models (not pretrained) on my native language texts, I was shocked by how rapidly they learn connotations. I find transformer-based LLMs to be very useful, yes, but not unsettlingly AGI-like, as I did before learning about them. I understand the usual way of building recommender systems, embeddings and things. Image models like Unets, GANs etc were very cool too, and when your own code produces that magical result, you see the power of pretraining + specialization. So yeah, idk what they do in interviews nowadays but I found my education very fruitful. It was how I felt when I first picked up programming.
Re the age of LLMs, it is precisely because LLMs will be ubiquitous I wanted to know how they work. I felt uncomfortable treating them as black boxes that you don't understand technically. Think about the people who don't know simple things about a web browser, like opening dev tools and printing the auth token or something. It's not great to be in that place.
Dude, thanks so much for this. I have a loud machine I’m using for ML stuff and I’ve been reluctant to move it to the basement as I didn’t want to have to run downstairs with a monitor for the odd issue. I also looked at KVMs and was annoyed at the prices. I have a R Pi sitting here I was looking to do a project with. This is perfect!
Can we talk about the waste that’s cast off when the reactor has finished its use of it? I understand the material will remain radioactive for centuries. Nobody wants that in their backyard.
Not to downplay the issue, but global warming is a much more urgent issue. It's expected to cause far more disasters a lot sooner than any potential disaster from our nuclear waste.
To answer your question more practically though, our nuclear waste would fit into a building the size of a football field. There's even a proposal to put it in some mine/cave in the western United States.
I would be pro-nuclear except for the waste problem. I don’t trust humanity to be able to confine the waste for centuries. For all we know it could be seen as some sort of “treasure” by future, ignorant, civilizations, dug up, and spread around in trinkets. If there is a way of permanently and completely removing them from the biosphere, I’d change my mind.
...if the alternative is climate change continuing to worsen until renewable technologies (particularly energy storage and transmission) are up to par and scale, which one do you prefer?
That's a tough call, but I'd prefer pushing harder on renewable research or finding a permanent solution to the nuclear waste problem, because, as bad as humans are at dealing with medium term consequences of our activities (see climate change), we have no track record of dealing with the long term, and ultra-long term consequences of what we do.
You have worse waste than nuclear waste, some chemical waste does not even degrade with time and will be there forever, doesn't that sound even scarier?
There’s a timing problem with this approach, when you get the money vs when the investor expects a return. The investor usually wants clear indications of what you’re doing with the cash to increase their investment. Holding it generally doesn’t sit well with investors as they could have given it to someone who would be doing things with it.
TaskRabbit | Fullstack React.js / Ruby Engineer | San Francisco | Onsite & Fully Remote | Visa considered | https://www.taskrabbit.com/careers
Join our mission to provide opportunities for millions of hard-working people across the United States, United Kingdom, and soon, around the world.
Our unique two-sided marketplace matches millions of customers with over 60,000 Taskers, offering convenience, quality and at accessible prices.
We’re in a period of explosive growth and have dozens of exciting job opportunities available right now at our San Francisco headquarters as well as being open to fully remote for some positions.
TaskRabbit | Fullstack React.js / Ruby Engineer | San Francisco | Onsite | Visa considered | https://www.taskrabbit.com/careers
Join our mission to provide opportunities for millions of hard-working people across the United States, United Kingdom, and soon, around the world.
Our unique two-sided marketplace matches millions of customers with over 60,000 Taskers, offering convenience, quality and at accessible prices.
We’re in a period of explosive growth and have dozens of exciting job opportunities available right now at our San Francisco headquarters as well as in New York, London, and Austin.
Every single day brings more chores and less time. Whether someone needs a handyman, house cleaner, mover or delivery person, TaskRabbit delights clients by finding the right skilled person in minutes. On the other side of the marketplace, Taskers earn a living by setting their own prices and defining their working hours. They have control to help people when and how they want in the most supportive marketplace.
Our goal is to revolutionize how people work. Our culture is collaborative, pragmatic, and dynamic. We're looking for talented, entrepreneurially-minded and analytical people who have a real passion for helping people do what they love—and have fun while on the journey.
TaskRabbit - http://taskrabbit.com/careers Senior Ruby / JS Engineer / React Native Engineers - San Francisco - Onsite - Full Time - Visa Transfers Available
Join our mission to provide opportunities for millions of hard-working people across the United States, United Kingdom, and soon, around the world.
Our unique two-sided marketplace matches millions of customers with over 60,000 Taskers, offering convenience and quality at accessible prices.
We’re in a period of explosive growth and have dozens of exciting job opportunities available at our San Francisco headquarters as well as in New York, London, and Austin.
We’re looking for experienced engineers comfortable working on the server in Ruby or on the client in JavaScript/React. For our mobile applications we're using React Native.
We believe strongly in modern development practices like code reviews, thorough unit testing, and smart collaboration.
reply