HNHacker News
TopNewBestAskShowJobs

javierluraschi

402 karma · joined March 23, 2014

Founder at hal9.com working on cloudish.ai

Previously RStudio, The Linux Foundation, Microsoft Research, and SAP.

x.com/javierluraschi github.com/javierluraschi linkedin.com/in/javierluraschi

submissionscomments
javierluraschi··on Ask HN: Why is ai.com redirecting to x.ai
Why would OpenAI sell that domain?
javierluraschi··on Ask HN: Why is ai.com redirecting to x.ai
I'm not asking about the company but rather the URL, ai.com used to be pointing to OpenAi now it points to Elon's x.ai company, why?
javierluraschi··on Ask HN: What AI tools do you use to boost productivity?
https://hal9.com for data analytics with generative AI, free when using CSV files.
javierluraschi··on Pandas AI
If you want to try something similar online, https://hal9.ai — Python based. Code is either processed directly on your database or in Pandas. Viewing the code is optional, but advanced users can opt-in to view and modify the code.
javierluraschi··on PaLM 2 Technical Report [pdf]
Seems like the file was in tabular format? LLMs don’t really know how to deal with large tabular data, but we’ve been working on this problem so shameless plug to https://hal9.ai
javierluraschi··on Tell HN: We should start to add “ai.txt” as we do for “robots.txt”
Related, there is also https://datatxt.org
javierluraschi··on Twitter is down due to buggy release?
Here is a screenshot: https://imgur.com/I5ws2oA
javierluraschi··on Twitter is down due to buggy release?
Same here!
javierluraschi··on CheatGPT
Gone through all the big-tech interviews, previously worked at Microsoft and a few more. I'd say that the only interview question I think survives here is what we've been doing at our startup: "For this role, here is a real four hour problem for you to work on, let us know when you want to start and we will send you the problem, send it back four hours later and then we will schedule some time to chat about it". If someone uses ChatGPT and solves it better than other candidates, that's great, good for them Don't think it's possible to solve it entirely with ChatGPT, but I could see some good candidates getting pointers to solve a 10%-30% of the problem to make them more efficient, before ChatGPT people could Google anyways and honestly I would expect they do.

I think a similar approach should apply to schools. Here is your math homework, solve it whichever way you want but then we will discuss it. I guess this would be the oral approach from the post but minimizing the time it takes to do the oral review. If your homework assignment has 20 questions and you discuss one with each student, you'll get a pretty good overview if they know what they are doing. If they want to ask ChatGPT for replies during this oral test, go for it, but they still get the same amount of time to discuss the question, is going to be really hard to get this done realtime. Also, I'm not sure ChatGPT really understand what it's doing to the point where an expert can prove it in limited time.

javierluraschi··on CheatGPT
Me to ChatGPT: reply to this interview question "What's the weirdest bug you debugged? What made it weird?" in first-person, make it a personal story.

ChatGPT: It was early on in my career as a software developer, and I was working on a web application that was designed to allow users to upload images and apply filters to them.

One day, a user reported a strange issue where some of the filters were producing unexpected results. Specifically, they were seeing images that appeared to be completely random and unrelated to the original image. At first, I assumed it was a problem with the filter algorithms, but as I dug deeper, I realized that the issue was actually caused by a very strange bug.

It turned out that the user's browser was caching a previously uploaded image and reusing it for subsequent filter operations, even if the user had selected a completely different image. This led to some very bizarre and unexpected results, as the filter algorithms were being applied to an entirely different image than what the user intended.

To fix the bug, I had to implement a workaround that forced the browser to always load the most recent version of the image from the server, rather than relying on its cache. It was definitely one of the weirdest bugs I've ever encountered, and it taught me the importance of always being vigilant when it comes to caching and browser behavior.

Me to ChatGPT: reply to this interview question "What's the weirdest bug you debugged? What made it weird?" in first-person, make it a personal story, make it using C++.

ChatGPT: I was working on a project that involved developing a custom compiler. One day, a user reported a very strange issue where the compiled program would crash whenever it encountered certain input files. At first, I assumed it was a problem with the input files themselves, but as I dug deeper, I realized that the issue was actually caused by a very strange bug in the compiler code.

It turned out that there was a subtle error in one of the class definitions that was causing memory corruption in certain circumstances. Specifically, the class had a virtual destructor, but the derived class failed to declare its own destructor as virtual. This led to undefined behavior when deleting objects of the derived class, which in turn caused the program to crash.

To fix the bug, I had to carefully analyze the affected class definitions and make sure that all derived classes properly declared their destructors as virtual. It was definitely one of the weirdest bugs I've ever encountered, and it taught me the importance of paying close attention to class hierarchies and virtual functions in C++.

Me to ChatGPT (since I haven't worked on a compiler): reply to this interview question "What's the weirdest bug you debugged? What made it weird?" in first-person, make it a personal story, make it using C++ about the time when I ported PyTorch from Python to R.

ChatGPT: One of the weirdest bugs I encountered was during my work on porting PyTorch from Python to R. One of the major challenges was ensuring that the R implementation matched the performance and functionality of the Python version. As I was going through the codebase, I noticed a particularly strange issue that was causing crashes in the R implementation.

The issue turned out to be related to differences in how Python and R handle memory allocation and deallocation. PyTorch relied heavily on dynamically allocated memory, but the R implementation was not properly managing this memory, leading to crashes and other unexpected behavior.

To fix the issue, I had to go through the entire codebase and carefully analyze all memory-related operations. This involved a lot of debugging and experimentation, as well as close collaboration with the rest of the development team.

Ultimately, we were able to resolve the issue and deliver a stable and performant port of PyTorch to R. It was definitely one of the weirdest bugs I've ever encountered, and it taught me the importance of paying close attention to memory management in C++ and other low-level languages.

Note: Likely ChatGPT can also generate a fake resume, etc. so yeah.

javierluraschi··on Show HN: Generating Streamlit Apps with GPT3
Hi there! We put together this package to generate Streamlit apps from prompts. The idea is to help you get over writer’s block. Let us know what you think!
javierluraschi··on Show HN: Sketch – AI code-writing assistant that understands data content
https://hal9.com is focused on building data apps with LLMs, would love to explore integrating and contributing to Sketch. If this sounds interesting I’m at javier at hal9.ai
javierluraschi··on Long Covid: major findings, mechanisms and recommendations
I was an early adopter of long COVID before it could be properly diagnosed, if I could ask for something, it would be to spread awareness for those who are still suffering.

The hardest part was, by far, feeling like I was making it all up, I felt a tremendous relief when I got diagnosed.

For me, it lasted almost a year, but I can say that I was lucky to fully recover.

There is a bit more info from my personal experience in this post: https://www.geekwire.com/2021/veteran-software-engineer-laun...

javierluraschi··on Show HN: Will data apps into existence with GPT-3
We are exploring using NLP for data app generation, specially as a way to bootstrap your project and not start from scratch. For those into building data apps, please give it a shot and let us know what you think!
javierluraschi··on Unpkg (NPM CDN) is having issues
Well, this is great, it's exposing all the libraries that are external but should be cached in our CDN.
javierluraschi··on Incident with Actions and Codespaces
10min troubleshoot time * 10K devs troubleshooting GitHub Actions right now... there goes one year for comunal dev productivity.
javierluraschi··on Introducing PyScript (summary of PyCon keynote)
We are not using libfortran not gdpr, we are basically using whatever libraries are available for the web. Since most data scientists don't want to use JS per se, you can build the apps as blocks in the Hal9 site or using a soon-to-be-released Python/R package, see https://notebooks.hal9.com

Feel free to check out our repo as well, all the "primitives" / blocks code is in the scripts folder: https://github.com/hal9ai/hal9ai

javierluraschi··on Introducing PyScript (summary of PyCon keynote)
At https://hal9.com, we built components for data science in native JavaScript to avoid the waiting times and download overhead if Pyodide. We found out the best tools for doing data science in the browser are a combination of Arquero and D3 and TensorFlow.js. At least for now.

We wrote our findings of this and many other libraries here: https://news.hal9.com/posts/data-science-with-javascript

javierluraschi··on 1 in 5 adults have post-Covid conditions
To the skeptics in here: As someone who experienced long COVID for almost 12 months, the first problem is that most people won’t even get diagnosed with long COVID.

I’m lucky to have had the time to research a lot on my own and find private healthcare to get to the bottom of this.

My initial diagnosis was stress and a second attempt allergies. Took me six months and tests to finally get a diagnosis.

Before I got a diagnosis, I was ashamed of telling people I had self-assessed myself with long COVID, it honestly sounded crazy at the time.

Now in some good news, I did fully recover after 12 months.

Rather than dismissing long COVID, I would ask that we at least accept we do not fully understand COVID. This is not to create fear, but to avoid dismissing suffering some are still experiencing.

javierluraschi··on Ask HN: Can you share websites that are pushing the utility of browsers forward?
https://hal9.com helps data scientists build faster web applications.

It uses WebGL and WebAssembly to process larger datasets, perform inference in the browser with TensorFlow.js, and enables running Python code with Pyodide.

javierluraschi··on Ask HN: What ML platform are you using?
If you want to build a web application on top of your ML project, give https://hal9.com a shot. We designed Hal9 with ease of use for deployment and maximum compatibility with web technologies that enable you to build ML apps with React, Vue, etc. We launched a couple months ago but could use some early feedback and users. Thank you!
javierluraschi··on WebGPU – All of the cores, none of the canvas
We have a few examples of time series data processing in Hal9 with WebGL (no support for WebGPU just yet):

- Simple regression with TensorFlow.js: https://hal9.com/hal9/historic-temperature-analysis - LSTM with TensorFlow.js: https://hal9.com/hal9/bitcoin-prediction - ARIMA with Pyodide.js: https://hal9.com/hal9/ai-fundamentals-arima

If you have question or would be open to considering using Hal9 to do time series analysis in JavaScript, let me know! You can find me in Twitter at https://twitter.com/javierluraschi

javierluraschi··on Show HN: Lovenotwar.js a minimalist banner to stand for peace
This Twitter thread across two software engineers from opposite sides of the world got us here; hopefully it encourages others to reach out, discuss and get closer in resolving their different world views <3

https://twitter.com/javierluraschi/status/150058723702029517...

javierluraschi··on Show HN: Lovenotwar.js a minimalist banner to stand for peace
I’ve been reflecting on how fast the situation in Ukraine is escalating, we have family and friends. I had to do something, even if it’s insignificant. The lovenotwar.js library adds a banner in support for peace and love, that’s it.

If anyone wants to help, please send PR.

Hoping for peaceful resolution. We all share this tiny planet, I hope we can de-escalate and coexist.

javierluraschi··on Pyflow – Visual and modular block programming in Python
We are working in https://hal9.com which is language agnostic and allows you to compose different programming languages; however, we are focused at the moment at 1D-graphs but have plans to support 2D-graphs in the coming weeks.

If you want a demo or just time to chat, I'm available at javier at hal9.ai.

javierluraschi··on Amazon and the “profitless business model” fallacy (2013)
I would recommend "The Innovator's Dilemma", explains really well why this happens.

In a nutshell, the argument is that successful companies fail to enter new markets because they have good management, not bad one.

In Ford's or IBM's case, a sane manager would not go invest in an unproven market, they rather maximize profits and growth on the market they are most effective at; however, this comes at a cost of missing new opportunities, which makes this a dilemma.

Successful companies could try to create little startups inside the big corporation; however, this rarely works out since you really need the entrepreneural spirit, align compensation to success and freedom without the restrictions of the parent company. The best alternative might be to create a VC arm that can serve later on as merger as acquisitions, but even then, is not guaranteed the startup would want to merge to the parent company nor that it would be the best decision.

javierluraschi··on Ask HN: How did front-end and back-end titles get introduced?
You remember when you started hearing about front-end and back-end roles? Early 2000's as well?
javierluraschi··on Ask HN: How did front-end and back-end titles get introduced?
Sounds like you are confirming that there was a time before the split, where it was more common for single person to run front-end and back-end, makes sense! I remember writing PHP at some point and running my own server and it wouldn't have crossed my mind to split the back-end work with someone else. I mean, at times a colleague would do the database, etc. so there was definitely specialization even back then, but it wasn't an explicit front-end vs back-end since we all had the skills to mess around with both.

However, I'm more interested to reconstruct how/when we started splitting into front-end / back-end roles explicitly. Any thoughts there?

javierluraschi··on Ask HN: What do you think about the no-code movement?
Shameless plug to https://hal9.com -- We believe there is a place for no-code, as long as you are not prevented from exporting or changing the components with code. We enable data analysis with parametrizable blocks which are 100% backed by open-source code, we believe you get the best of both worlds with hybrid solutions like this one.
javierluraschi··on Show HN: Hal9 – Low-Code Analytics for the Web
Here is a link to a live Q&A stream we had as well: https://www.youtube.com/watch?v=A93003kB4K8&t=270s
← PreviousPage 3 of 7Next →