erikgaas

I've been reading the book. The important part is in 2004 they were offered to be acquired by serono. Because of their governance they were able to say no to the acquisition. Meanwhile serono was later acquired by merck and largely shut down. Back in 2004 novo was about $2.5 a share. Now at its current price even after a huge downturn it is $43.3 So that 17x gain you can attribute advantages with how they governed themselves at that time. Is it a silver bullet or always make stock go up and to the right? No of course not. Are they making governance mistakes now? I'm not sure. But most companies probably would have sold for a ton of cash and if so everyone would have missed out on huge gains, not just in money but in technology and public good.

One thing I loved about the first solveit course was how create the community is. It goes back to fast.ai too, but everyone is super kind, smart, and has diverse backgrounds.

We've captured a slice of that on our main site. Testimonials: https://solve.it.com/testimonials Some blog posts: https://solve.it.com/#showcases on the main page

And on of the students even made a project dashboard page showcasing all the things everyone has built! https://solveit-project-showcase.pla.sh/

He even blogged about it : ) https://himalayanhacker.substack.com/p/how-i-built-solve-it-...

Here is a video showing off using solveit for creating a web app. https://youtu.be/DgPr3HVp0eg?t=3120 To reiterate other comments, this is more about the methodology than the tool, but it is fun to see the tool in action too.

Right. I agree, but I think you are appealing to generosity when it works just as well if you appeal to greed and selfishness.

If I'm a parent who does not intend to take advantage of the program and therefore not to get any benefit directly, and I assume the program is done well and not rushed, I could reasonably expect:

- More parents able to be in the work force (immediately) - Better metrics for the young children entering. Especially for at risk. - Savings from less crime in the future. - Higher attainment of students when they enter the work force later. - Higher birth rate??? (probably not but this one is interesting regardless)

My understanding so far is that this leads to spending savings in addition to QOL of life improvements. And that's just for me. I want to live with less crime and less tax liability.

Asking for additional waivers imo just increases the cost in areas that will not as directly achieve the benefits of the program as stated. The only reason to ask for it is as a negotiation tactic.

I think the most important thing is to focus on the quality of the program and make sure the resources are there. And to make sure opportunities persist to prevent "fade out". I think that might have been the difference between Oklahoma's success in pre-k vs a program in Tennessee.

Templating is great. I think I prefer fasthtml because I can apply all of the python tricks I know. Python arguments seem to work very well. Having a list of dom elements and splatting them into another component would feel very similar to templating.

Imo they are abstractions but not quite as bad as you / myself / everyone else is used to. Monsterui is moreso an extension of fasthtml. It doesn't hide any of the underlying API. Same with fasthtml with respect to htmx.

Htmx also may leverage js but it is meant to patch http functionality based on how http "should" function. See the hypermedia book for that discussion. You don't need to really know that it used js. You don't interact with it via js, just the dom.

As for starlette I'm not currently aware of any server stack that doesn't have some convenience library for that.

My point is that the frameworks which makes us very pessimistic every time a new approach comes out frequently try to make something really easy but once you get far enough you realize that you have to break their abstractions. And create hacks just to access the lower level implementation. I think you'll find something like this very different. When you need control you'll find everything to be much easier to decompose such that you can operate at your needed level of abstraction.

There is no reason for the author to be misleading about Table as a "high level component". There are more than enough examples that fit your definition. Look at the code for dashboard and music and you will see plenty of abstractions on top of the html components. https://monsterui.answer.ai/music/

If you want a custom Div you can just define a function for it. The definition of DivLAligned is just for convenience. You can see its implementation is fairly trivial.

https://github.com/AnswerDotAI/MonsterUI/blob/main/monsterui...

Much better really? Looking at the muni map, pretty much all transit clusters around downtown. And bart doesn't even stop at 4th and King. You have to go to Millbrae for a close-ish transfer. Added to that the transit center is very close to the ferry building. And then you have golden gate transit. And it's a hub that has tons of space to accommodate a ton of transfers.

https://www.sfmta.com/maps/muni-service-map

The other important implication is how Caltrain could be routed through underground tunnels to Salesforce transit center. 4th and king isn't super nice and, correct me if I'm wrong, but it isn't well connected to other modes of transport. Maybe some muni buses.

I took a waymo in Chandler that had no safety driver and it was able to make unprotected left turns just fine.

I work for Lyft's self driving car division, Level 5! Nbdev has been great. I use it a lot. Thank you for all of the work you've put into it!

So I use this in production at my company. It's an awesome tool. Personally when I'm coding in python I like to prototype in jupyter, copy code over, and then reimport anyway. Nbdev streamlines everything so I can write docs, tests, and code all in one place. And since the docs are just a jekyll site I can copy it to our documentation aws bucket in continuous integration. And with one command I can run all the notebook tests in CI as well.

The packaging is also really well thought out. I don't have to stress out about connecting setup.py with whatever publishing system we have. The settings.ini makes things sane and I can bump the version whenever I want.

A get a lot of skeptical looks when I say the source code is in notebooks, but that's just syntactic sugar for the raw source code. You still get to edit the raw code files and with one command sync everything with the notebooks. From my point of you it is close to a pareto improvement over traditional python library development.

I haven't used it in a while, but I remember having good experiences with Sumatra.

https://www.sumatrapdfreader.org/

Been using this for work projects. A lot of raised eyebrows when people hear jupyter first development, but the automated docs, flexibility with prose, inline testing, out of the box pip packaging, and git integration make it well worth it. A bit of a learning curve, but very rewarding.

I took a neuroscience course back in the day that discussed findings related to autism. I don't remember everything, but there was a discussion on excess dopamine early in the brain and a later depression of glutamate levels. All in all this would lead to more sparsity of neural activations in the brain. This made sense. Too much dopamine would lead to excess plasticity, causing a normal amount of glutamate to have too great of an effect, so the brain compensates by producing less glutamate, producing the same amount of neural activations but now sparsely spread out in a post-plastic brain. From my naive common sense perspective, it would make sense that an autism prevention would be some mechanism to limit plasticity during development. I'm really stretching argument here, but maybe this has some connection to autism brains "kicking in" early.

The connection to estrogen is an interesting one. Estrogen does have an important effect on neurotransmitter receptors, including dopaminergic, but my speculation is worthless here. Is there anyone here that has a better understanding of this that can provide some clarity?

Sometimes it is useful to make useless things. It will build experience while being fun and making you more able to come up with "useful" things. Tried this with a friend of mine. Our plan was.

1. Take webcam at desk. 2. Build model that detects whether he himself is sitting at his desk. 3. If it is not him, you've detected an intruder, so spray the intruder with a squirt gun or something.

We only got as far as a model that distinguished between him and other people, even deploying it according to this guide. https://course.fast.ai/deployment_render.html

Also the material is just really interesting. If you want to know how a lot of products work, this is one of the most fun ways to learn.

You can but it's not ideal. What if you want to write your own cuda kernel for your experiment? Python isn't really setup for this easily unless you want to throw odd c++ integrations into your code. Swift is designed to be a direct match to the underlying instructions. This would make deep learning much more expressive and flexible in Swift, with less errors.

This question is addressed extensively in the course. Check out the last two lectures, they do a great job of going over lots of different reasons.

I was lucky enough to participate in the lectures earlier this year. I've watched every fastai course before this and I will continue to watch them, even the introductory material.

That being said, this course is completely different than anything done before. This course takes a look at the very most basic functionality of neural network libraries and proceeds to implement the fastai library totally from scratch. So it's really a great deep learning + tooling + api design + programming language + research implementation course all in one.

The last two lectures on Swift were especially fun. Python isn't great for deep learning because you always have to live at arms length from the implementation through a c++ wrapper. Swift can compile straight to the low level system commands. LLVM, MLIR. Still learning this so I'm not the best person to answer questions about this specifically. :)

So Swift has this awesome feature set of being well designed, pretty similar to Python, statically typed, and low level. These lectures are hosted by Jeremy Howard and Chris Lattner. Sometimes the most valuable learning I can do is to just listen to the banter between two exceptionally experienced experts, so this make the course that much more enjoyable.

Five starts would recommend. Get the t-shirt.

FWIW I use the %debug magic command in Jupyter and it has been a great experience. I'm pretty ignorant of the enterprise debugging tools so take that with a grain of salt.

I just did this in my University lab as well. Most people aren't savvy with Linux, so having normal accounts with Jupyter port forwarding is out of the question. JupyterHub is just about the lowest friction I can possibly make it for introducing the Python data science stack to non data scientists.

Pets, cats, and dogs only live until they're 20. AND pets DO have lots of problems with dental health. What is with all these people questioning flossing here? Floss your damn teeth. Worst that can happen is the poor guy you work with doesn't have to smell the decomposing food coming out of your mouth. JFC

I was a student in neuroscience, and I got the same impression from people in the field that AGI isn't close. However, arguments were typically from the standpoint of whole brain simulation. We know very little about the brain. And we know computer scientists know less about the brain than the neuroscientists, so how could we possibly be close to replicating that? There would probably be more progress if CSCI and Neuro would communicate more. I don't think the neuro people appreciate the opportunities in the hardware and algorithm space, while the CSCI people don't typically study neuroscience, so AI hugs this interesting intermediate space where it only looks like neuroscience if you squint a lot. Some people think that we need to go all the way to simulating ion channels. I think this is probably silly and we can abstract better than this. In any case you are going to see a lot of disagreements just because of where people want to draw the line for biological fidelity.

AI developments have been phenomenal in the past few years. And the economic return makes me expect that this race will continue faster and faster. I don't think human brain project criticisms make this any less of a reality. Even now it is hard to find a well-defined task that can't be performed better by a computer than a human. Humans are really good at dealing with ambiguity though. So a robot might do better driving on well defined roads with nice lane boundaries, but humans are good at dealing with construction, or negotiating between difficult drivers.

We have already been able to generalize just about any modality you can think of to be processed by neural nets, and sometimes at the same time. If you squint this feels almost like different regions of the brain. (Vision, hearing, speech) But I have reservations about anthropomorphism since it can cause arguments that keep people from just making something that works.

If you think Kurzweil's predictions are a fiction, you are probably right. But I think that's mostly because predictions on those scales are very sensitive to interpretation.

For me, I think the future according to my perception of what Kurzweil is saying will probably be way different than reality. But the future of AI will probably have an equivalent impact and be just as surprising as if my perceptions were accurate.

Thanks much! It should be working now.

You don't really have to know anything. And the current free course does not have anything to do with neural nets. They use monte carlo, kalman filters, particle filters, PID, etc. You should feel comfortable with probability theory and if you know some linear algebra, the kalman filter example is going to make a lot more sense.

Yes those are all important parts in the development of cancer. But you have to take into account the tools that cancer has to take advantage of when it comes to invasiveness and proliferation. DNA instability is the root cause which results in incorrect signalling, cell cycle progression, cell morphology, and immune system evasion. But take angiogenesis, an extremely important tool for a developing embryo, and how ubiquitous it is for a growing tumor. But we can go farther and consider the factors that influence the guidance of a cell during development. Accutane or Isotretinoin is very similar in structure to retinoic acid, a guidance molecule in the posterior embryo. Consequently isotretinoin causes birth defects. Interestingly isoretinoin has also been implicated as an anti-cancer treatment. Thalidomide, which also interrupts proper development has also been used as an anti-cancer treatment. My point isn't to dispute the importance of DNA damage and gene expression profile changes, but rather to emphasize the importance of metastasis and its apparent parallels with embryonic development since metastasis is the main reason (sans leukemias) why cancer has such a high lethality.

One thing to be cognizant about is that cancer is really a disease caused by the reverse aging of particular cells. If you make a cell think it is too young it is going to become invasive. We don't normally think of embryonic development requiring invasive behavior of cells, but that is certainly the case. Take a look at the development of the cerebral cortex and how cells must travel through other layers to reach their destination. It appears that they are targeting telomerase which is very interesting and parallels cancer where its over-expression is used often as a marker. But then again telomerase under-expression can occur in early-stage cancer and is thought to contribute to genome instability/mutation to affect other genes before it is eventually over-expressed. So this kind of experiment is really exciting and will almost certainly give us insight into many different medical fields. I just hope that her experiment winds up on the preventative side of cancer rather than the causative one.