I listened to you on the latentspacepod the other day and tried this method myself.
1. Used GPT-4 to build functions that pulled in sample datasets from my BigQuery.
2. Used GPT-4 to insert them into a sqlite database.
3. Upload the zip file into Code Interpreter.
4. Now I'm in an environment to quickly self-iterate on GPT-4 generated SQL.
It felt incredibly powerful. Previously I was developing with the latency of 100Gb+ tables – now the sqlite db responds instantly, and I don't even need to write the code.
I’ve used this library for a number of projects and it’s a joy to use. I don’t think it’s an understatement to say it’s paradigm shifting – to the extent that once you have your environment set up, you are free to code, think, iterate, deploy and document your projects all at 99% of the speed of thought.
There seems to be a lot of discussion in here around the pitfalls of jupyter, and notebooks, and the poor coding practices of data scientists. If you haven’t read the article or used the software I’d like to highlight that all of these (legitimate) complaints are exactly what nbdev2 was created to address, and in my opinion very successfully solves.
The way it works is that everything runs off a master notebook, and then with one command: libraries are built, git diffs are magically fixed, tests are run, documentation is automatically created. It doesn’t fundamentally change your workflow in any way, it just abstracts and automates away all of these pain points.
There’s a reason that everyone uses jupyter notebooks. They are fun to use, they are great for exploring and developing ideas. And (minus the aforementioned git collaboration issues) they are great for sharing with others, which is a huge part of the wider data science ecosystem. We don’t need to recommend avoiding notebooks, and allege they are just for beginners. We need to use tooling which addresses some of these final issues with writing mature software. And I'd like to thank the authors of nbdev for this.
The people who look down their noses at notebooks can continue to do so – but what they will find is that nbdev quite effortlessly leap-frogs over these sneered complaints, and allows you to write better software more productively.
I think you’ve done a good job of summarizing the landscape – it matches my experience.
The progression seems to be the decline of the nerd-nerd monoculture in the professional tech environment. With some irony, the increase in diversity is from the non-neurodiverse. Product managers, engineering managers, vertically integrated teams.
I believe on the most part this creates healthier productivity for the business – but likely at the social expense of the neurodiverse. Their technical skills can be more readily exploited by the hustlers.
An industrial designer would sketch the outline then the images would be brought into parametric CAD software as references. The 2D image can be built into 2D/3D lines, then these curves are used to create surfaces which finally are stitched together. The nature of complex surfaces interacting and joining seamlessly is where the difficulty lies. "A class" surfacing is an entire speciality and can take a career to master.
https://en.wikipedia.org/wiki/Class_A_surface
Thanks for this. I'm curious to hear why something that has such remarkable potential to improve lives, has somehow been channeled into a social taboo.
So on the surface of the shell planet, the effects of gravity would be the same as normal, but when you're on the inside there is no net gravitational effect? ie. you'd just float near the inside surface? That doesn't seem intuitive. Wouldn't the force from gravity be acting towards the centre of mass in both cases?
Edit: Actually never mind, I think the density of the shell is what threw me off! On the outside the net gravitational effect is the local shell (RHS) plus the rest of the shell (RHS), but just on the inside - the net effect is the remaining shell surface (RHS) minus the local shell (LHS). Because the effect of gravity is 1/r^2 everything works out!
While the author makes a good point about how humans will never leave the solar system, there are a few possibilities to continue human legacy. Ethics aside, sending seeded capsules to a habitable exoplanet could one day (millions of years) evolve into an intelligent species. Although not human, a DNA signature or some other artifact could be engineered to validate it as a human colonization.
Additionally you could explore the possibilities of fleets of nano sized probes, which over the course of thousands of years, confirm the habitality of an exoplanet, and build a crude nursery for sperm (which could be sent at a later date). This method makes the energy/momentum problems slightly less impossible.
It felt incredibly powerful. Previously I was developing with the latency of 100Gb+ tables – now the sqlite db responds instantly, and I don't even need to write the code.