gabelschlager

6 ms·

1.
by gabelschlager 1mo ago
Is it? Especially if you go the K8S route, setting up a runner infrastructure can be done within a day. Based on my experience, there isn't a lot of maintenance involved. It's mostly a smooth experience.
2.
by gabelschlager 1y ago
Smartphone is required for everything there, yes. Signing up for services, authenticating yourself (e.g. when entering a train station), payment, social media, etc. Computers used to be expensive and people had less money back then, so most
3.
by gabelschlager 1y ago
I think the endgame is a shift toward a platform of services that tightly bind users to a single LLM provider. Right now, many small startups are essentially just thin wrappers around ChatGPT. Once it becomes clear which ideas and solutions
4.
by gabelschlager 1y ago
According to Hiro's latest update on the matter, no. [1] https://x.com/hirox246/status/1912603340292448258
5.
by gabelschlager 2y ago
>Zelensky implied Russia would invade the U.S. Do you think that’s likely? He did not. What he did say is, that should Russia start to expand the war to other countries (as is very likely if Ukraine falls), these effects will be felt by
6.
by gabelschlager 2y ago
Honestly, Regex is not nearly as complicated as most people make it out to be. Read through the rules and then design some simple expressions.
7.
by gabelschlager 2y ago
I quite liked ReadEra. Scans your phone for all epubs, allows grouping and organizing them, but most importantly, has a good looking, customizable reading interface.
8.
by gabelschlager 2y ago
I think they are best at information extraction/classification tasks, especially for complex tasks with little to no training data, and data synthesis tasks. However, you should always test if simpler models can already perform the tas
9.
by gabelschlager 3y ago
Just talking about the software-side of things: I think they actually have been adding very neat features with the past few updates. Automatic image OCR, so you can copy/paste text, inserting text via camera, the transformer-based keyb
10.
by gabelschlager 3y ago
Even you are still personifying the model. There is no motivation. An input sentence is multiplied by a series of weights and the resulting output vector transformed back into a text. At the end of the day, it's a program like any othe
11.
by gabelschlager 3y ago
Most of the time I try it to use it, the website won't allow me to use the '+' in an email or will never send the verification mail to activate my account.
12.
by gabelschlager 4y ago
I mean, even here people regularly don't understand what ChatGPT is and what it can do. I wouldn't expect reddit to be any better.
13.
by gabelschlager 4y ago
Well, Chomsky already dismissed corpus based linguistics in the 90s and 2000s, because a corpus (large collection of text documents, e.g., newspaper, blog post, literature or everything mixed together) is never a good enough approximation o
14.
by gabelschlager 4y ago
Due to upcoming EU regulations they most likely have to update their policy.
15.
by gabelschlager 4y ago
It's a language model trained to return the most likely answer given a specific prompt. The goal of the answer is, to simulate a conversation and to sound realistic. ChatGPT is not grounded in reality, it does not know where its "
16.
by gabelschlager 4y ago
Is it really that much of a problem though? I own PC games that are close to 25 years old at this point and have problems installing and playing them on the PC I build two years ago. I have hard drives and USB sticks that are from 2005 and
17.
by gabelschlager 4y ago
I had a similar problem to you. While it was easy for me to complete assignments on Udemy or similar, I didn't develop a deep understanding of anything I am doing. What finally helped was doing the free Java course from the University
18.
by gabelschlager 4y ago
What do you normally use a search engine for? Sure, if it's just specific, factual questions, AI generators might become good alternatives for that in the coming years. However, currently they hallucinate too strongly to be trustworthy
19.
by gabelschlager 4y ago
There are sites like japanese.io, that allow you to copy/paste any text and get vocabulary lookup, furigana support, allow to add words to a flashcard deck and other things. I really like their idea and feel such a platform would add a
20.
by gabelschlager 4y ago
You say that your semantic search works best with short texts? What exactly do you define as a short text? Just single sentences, paragraphes or even a couple of pages? Other than that, I agree with the other commenter that more complex dem
21.
by gabelschlager 4y ago
Any of the courses from the University of Helsinki [1]. They are free for everyone, have good explanations, and most importantly, great exercises to practice programming with an auto-grading system that gives feedback when you make mistakes
22.
by gabelschlager 4y ago
It's good to keep in mind that games like League of Legends alone have over 100 million monthly players. While I agree that mobile is far bigger (especially thanks to Asian countries), PC's slice is too big to ignore. And just goi
23.
by gabelschlager 4y ago
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law. I usually take my prediction, multiply the time by 2 and add another 1/4 of the time to it. If it turns out to be fas
24.
by gabelschlager 4y ago
The nice part about Germany is, that you don't pay any tuition fee there (except Baden-Württemberg). So doing a Master's there is a lot cheaper than in the US. As for CS, it is possible to work there with a BA, but generally speak
25.
by gabelschlager 4y ago
Btw, not having any sort of contact information (address and name of the business/person running the site) is illegal in the EU. And generally just highly unprofessional.
26.
by gabelschlager 4y ago
There has been a couple of studies on the topic that observed network traffic and never were able to found anything. So despite the persistence of the rumors, so far this remains a conspiracy theory.
27.
by gabelschlager 4y ago
Try building outside WSL. I had the same problem, but moving the app to standard Windows improved the startup times from ~30s down to ~10s. Never bothered figuring out what caused the differences.
28.
by gabelschlager 4y ago
I didn't build it myself, but I maintained one for a while as part of a student job. The website itself was hosting various resources for researchers, etc. Each resource was annotated via a specific XML format and XSLT was used to dyna
29.
by gabelschlager 4y ago
Yeah, the Semantic Web at least tried to be actually useful!
30.
by gabelschlager 4y ago
I think real research in this regard is still rare. For one, because not many software solutions exists. But more importantly, performing research studies in schools tend to be difficult (as parents, teachers, the school, etc. have to agree

More ›