oh_teh_meows

Great work! Some questions:

1. How would you compare Stacker to Retool?

2. Would the data from Stripe/Intercom appear as just another sheet/table?

3. One thing I learned that, despite the promises of no code solution, a non-technical user would still not use them, because they lack the know-how to model their domain knowledge in a no code tool. Have you run into these users? How would you work with them?

The Code component seems like a catch all. If it were to support actual AST, surely

<Code>const a = 3</Code>

Would have looked something like:

<Assign> <Lhs constant><Id>a</Id></Lhs> <Rhs><Literal>3</Literal></Rhs> </Assign>

Author here!

What can you do with it:

1. String together Unix-like text processing functions into a text processing pipeline.

2. Share said pipeline with friends and co-workers through a transparent link.

3. Since the functions are encoded in the URL, you can even compose functions without getting on the website! Just start typing a link: www.textpiper.com/grep?total/col?2&3/sort?asc/replace?dollars=yens

4. Your data is safe since it doesn't get out of the browser, but I'm aware that composed URLs could get logged. In future I'll add opaque URL.

It's very much a prototype and a work in progress, and as you can tell already, my UX skills leave a lot to be desired. Feedback is very much appreciated!

Really really cool! I wonder if we can do something similar with programming. Suppose we have a conceptual framework where each module/system is a widget made up of slightly simpler and well-defined widgets. If we can specify what we want, perhaps a system can be devised that would try to build it from a database of widgets. If some of the constituent widgets are not found, the system could recursively specify each one and find/construct.

In this recursive process, the language used to specify the requirements of a widget can be a changing DSL whose grammar and basic constructs co-evolve with the complexity/abstraction level of the widgets.

This seems feasible because we (software engineers) are one such system. Starting from basic transistors, we build ever higher abstraction layers along with the language used to specify them (circuit diagrams -> microcodes -> assembly -> C -> DSLs). I believe multi-layer neural networks are also a prime example of such a system.

you're right about that, so I think what it could help us with is to provide a sort of guidance when we have a particular end-goal in mind.

This might be a little off-topic, but I wonder if the idea of nano-passes can be applied to other areas. For example, can we predict the future of technology by gradually evolving their capabilities? Or say start from some point in the future (interstellar travel!), and gradually work backwards what we think is needed?

Or how about predicting geopolitics/economics?

To use this method effectively, I think we'll need to fully specify, to the extent possible, the condition of each stage in time. That way, at any stage, we can see what constituents might possibly interact with each other and evolve something new.

So...eat cheese the night of your study cram? (Bad idea if you're lactose intolerant)

Thank you!

I had come from a Compilers/computer vision and distributed systems background, and just recently I decided I'd quit my job and bootstrap my own little SaaS business. My web development knowledge had basically been stuck in the era of PHP and XmlHttpRequest, so I was in for a huge surprise when I decided to dive into modern day web development. For 3 months I couldn't do anything useful; it's just straight up a lot of fumbling in the dark with lots of alternatives and opinions on how to do something. The most aggravating things I found were the proliferation of tutorials written by beginners, and libraries/frameworks with lots of new terminologies which are just rebranding of existing CS concepts. It's aggravating to me because I made a distinction between knowledge from sciences and knowledge of made up words that won't be relevant in a few years (e.g action reducers, action creators, state store, etc), but in the end I decided to just suck it up and internalize all these words so I can move on with my project. They were useful to someone's project, so surely I could humble myself and learn something from them.

This piece may seem like a satire on the current state of javascript's ecosystem, but as far as I'm concerned it's an accurate portrayal of the pains that one goes through when picking up front-end development from scratch. I can't reproduce my initial development environment from memory even if I wanted to (too many npm dependencies and command line incantations). I've resorted to freezing them in a git repo and forking from it when I want to start a new project. I hope in time I will figure out the bare basics for what I plan to do.

Reflecting on my learning experiences, I think perhaps I would have progressed faster if I had an opinionated mentor who could hide from me the diversity of the ecosystem, or a guide if you will, until I'm ready. Entering web front-end development is just like entering a bazaar with thousands of sellers calling out to you. You wanted to buy a simple woodwork table so you could get on with your home improvement project, but you left the place 8 hours later with a bunch of tools that would be used to build said table instead, and you're nowhere close to getting started on your actual project.

This maybe the beauty of the ecosystem, but getting into it wasn't a pleasant experience for me. Now that I've gone through that phase and can get started on the real work, I hope to spare others the pain with what I've learned.

I understand that the achievement in question was made in a high pressure environment; but if we get room temperature superconductors (at standard atm pressure), does that mean large scale hover cars/boards deployment will be viable as well [0]?

[0] https://youtu.be/PXHczjOg06w

@ML experts: could a model be trained from ECoG data so that we can better isolate signals from EEG?

I wonder if the state that C++ finds itself now is a result of sub-optimal compromises between different interests in the standards committee. Year long arguments of what ought to be included or how they ought to be done may cause folks to lose sight of what's important, you see.

I still like C++ though. At least a subset of it anyway.

Look again. The downward trend of Java was arrested and it stayed on a plateau when Android began becoming popular; then come 2015 they both came down together.

If we're using Google Trends as a proxy for what happened, then I'd say there's some truth to Android saving Java from its decline.

Does anyone know of a similar tool but for programming/connecting to a remote terminal? Electronic ink display with week or even month long battery life? CPU doesn't need to be super fast, just enough to do some fun coding would be sufficient :]

This might be a contentious question, but how did it come to this in US (likely not the only country suffering from problems with law enforcers), i.e best course of action would be to not talk to the police? I imagine if I were in say, Japan, I wouldn't have to worry about cops having it out for me (especially if I'm innocent).

OP's probably referring to induced demand [0], or something similar to Braess's paradox [1].

But I see your point though. If people are incentivized to share a car then that might reduce the load.

That said, my suspicion is, once you take the driver out of the equation, you can do so much more than just ferrying people. E.g courier/shipping services, or mobile food stalls (food truck that sells food while moving on the highway, once engineers figure out how to safely and autonomously speed match two cars), etc. The induced demand may be greater than the savings from ride-sharing.

[0] https://en.wikipedia.org/wiki/Induced_demand [1] https://en.wikipedia.org/wiki/Braess%27s_paradox

That video reminds me of a comic I sketched out years ago. It's about this rag tag armed group of hackers who specialize in intelligence espionage in the near future. There's a payload specialist, security specialist, data structure specialist, etc. Much of the drama unfolds in VR space where the hackers can be seen frantically querying/manipulating data structures directly by hand whilst evading detection. It's supposed to be educational as well, explaining CS topics through stories. Think the movie hackers plus GitS, but with attention to accurate portrayal of CS knowledge. It's basically my vision of what the future could be like.

"A technique called weight quantization, for example, represents each neural network parameter with only a few bits, sometimes a single bit, instead of the standard 32...The models are equally accurate, but the compressed version runs about 20 times faster."

This is pretty great. How can we tell if a particular ML problem is amenable to weight quantization without sacrificing accuracy?

Is this about the rising cost of housing? My understanding is that it has more to do with nimbyism and proposition 13 than just engineering wages pricing out locals. That said, I don't know about the situation in Canada, so maybe someone can shed some light.

Sweet! Thanks I'll check it out.

> The video in the article shows basically shellcode injection, but that's not timing-sensitive, it'd just take longer for a human. And, as seen above, similar things are possible for humans, if just less convenient to do so.

Wow that just gave me an idea for a story. Humans found out they are mere pawns in a simulated reality, discovered a 'hack' to alter reality, but the 'hack' would take hundreds of years to complete. So generations of humans toiled over at completing the 'hack', passing the baton through each generation.

There could be so many possible storylines from this - corruption/destruction of reality, dictator wanting to change the past, or a cult hell-bent on changing reality, or sorcerers practicing 'magic', or a lone protagonist who's on the verge of completing that hack after hundreds of years, but suffer from ethical conflict and existential crisis.

Theory of relativity would say that there are many 'nows', which one you're in depends on which frame of reference you're in :)

To a photon traveling at c, the birth, growth, and death of the universe happen simultaneously; so to a photon, there is no now. From its perspective, it dies just as quickly as it comes into existence.

> Is there not just one ever present now in which all thing change?

Sure there may be; but going by the hypothetical possibility I raised, there really isn't a now. The passage of time could be an illusion due to your memory being in a particular unchanging state that has existed 'forever'. Who's to say the 'now' that has just passed wasn't just you remembering it?

As the universe approaches heat death, quantum mechanics dictates that particles jump between states randomly, such that passage of time cannot be measured meaningfully.

Or that there really is only one unchanging state, and what you perceive to be the passage of time is really just a very particular configuration of your 'memory' through which you 'remember' a different state, and that hasn't really changed at all.

Hm not necessarily so. Some people could struggle to understand a concept even with all the information available, and could really use some tutoring from a professional. That's when you start to notice the difference between just getting information from an article, and receiving wisdom from an expert. This service could certainly help with that.

Not in the traditional individual sense, I think. All AI drivers across the whole world will seamlessly exchange and synchronize with each other their learning experiences, resulting in self-driving vehicles that improve at an exponential rate.

It's basically an AI collective, whose 'brain' scatters across data centers around the world.

I used to play with roborealm when it was still free; it provides the same sort of high level abstraction to experimenting with computer vision and comes with a big library of filters/kernels.

How would one go about verifying that the debris comes from the missing plane, assuming plane-specific serial number is not found? Any expert care to shed some light on this?

What are the advantages of ZeroDB over a database that uses homomorphic encryption like CryptDB?

Do they cater to different uses?

Direct brain interface requires invasive surgery and is still a ways to go. Its constant need for recalibration basically precludes it from any useful utilization, much less commercialization of it on a massive scale.

Even if the AI part is mature, I do not think speaking at length (in a public setting) telling your computer what to do is desirable.

I think our best bet is a device that reads neural signals going to our voice box. We can "speak" without vocalization and still have the device pick up our words. I think they call it EEG sensing of imagined speech.