Egress is one of the most important concepts to understand when dealing with infrastructure providers like AWS, yet, so many people aren't aware of it until they receive their bill! AWS should make egress billing front-and-center in the free tier setup so that developers aren't falsely lured into thinking that a free tier service is actually free for a given period.

0
0
0

Wait, you mean to tell me you don't sort your imports by category (default, named, named with type annotation) and then alphabetically within those categories? How do you sleep at night? Standardizing import declarations this way makes scanning them a breeze and reduces cognitive load. You should try it.

0
0
0

If it's true that you're the average of the five people you spend the most time around, mine would be: spouse, two kids, ChatGPT-4, and Claude 2. (Cats notwithstanding.)

0
0
0

Entirely too much of my time is spent using Block Contact & Move to Junk.

0
0
0

I look forward to the day when I can have a LLM analyze a GitHub repo and have it generate all of my tests/documentation. That will be a good day.

0
0
0
https://cookbook.openai.com

The OpenAI Cookbook is an open-source collection of examples and guides for building with the OpenAI API.

0
0
0

Recently, I saw a smart targeted marketing tactic: sending automated pings to the websites of users in your target demo. If they use advanced analytics, they'll probably get curious and look up your company via the ASN org field. Experienced this today—kudos to Scaleway!

It's a win-win: boosts the target's vanity metrics and is cheaper than regular online ads.

0
0
0

Mailgun's user onboarding process is infinitely better than SendGrid's—with the exception of their tricky sign-up form which leads you to believe you have to add your credit card to start your trial. Almost perfect!

0
0
0

It's annoying when the most commonly used software shortcuts (in cases where the user's right hand is likely on the mouse and their left hand is firing off shortcuts) requires the user to physically move their left hand to the right side of the keyboard. Sigh. Are designers not analyzing the most often used shortcuts and then deciding which keys to used based on actual user behavior?!

0
0
0

It'd be funny if the concept of jinxing actually had a scientifically grounded base in reality related to observability (e.g., the double-slit experiment).

0
0
0

Sometimes I'm so deep in my work that I fall off the social media map for months at a time. On the other hand, it's an incredible feeling to be able to hit so many of my annual goals in a single quarter. I feel like my learning rate is exceedingly exponential—a big part of which is due to my daily interactions with LLMs. Also feeling a bit like Johnny Mnemonic when his brain is at max capacity.

0
0
0
https://www.youtube.com/watch?v=NUhrF0xkhhc

Add this to my list of things I don't need that now I really need.

0
0
0
https://www.youtube.com/watch?v=oLiheMQayNE

E11: OpenAI's GPT-4 Discussion with Red Teamer Nathan Labenz and Erik Torenberg.

Very important discussion, especially starting at 28 minutes in, re: lessons learned by one of the red team members.

0
0
0
https://arxiv.org/abs/2303.12712

Sparks of Artificial General Intelligence: Early experiments with GPT-4

Abstract: "Artificial intelligence (AI) researchers have been developing and refining large language models (LLMs) that exhibit remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition. The latest model developed by OpenAI, GPT-4, was trained using an unprecedented scale of compute and data. In this paper, we report on our investigation of an early version of GPT-4, when it was still in active development by OpenAI. We contend that (this early version of) GPT-4 is part of a new cohort of LLMs (along with ChatGPT and Google's PaLM for example) that exhibit more general intelligence than previous AI models. We discuss the rising capabilities and implications of these models. We demonstrate that, beyond its mastery of language, GPT-4 can solve novel and difficult tasks that span mathematics, coding, vision, medicine, law, psychology and more, without needing any special prompting. Moreover, in all of these tasks, GPT-4's performance is strikingly close to human-level performance, and often vastly surpasses prior models such as ChatGPT. Given the breadth and depth of GPT-4's capabilities, we believe that it could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system. In our exploration of GPT-4, we put special emphasis on discovering its limitations, and we discuss the challenges ahead for advancing towards deeper and more comprehensive versions of AGI, including the possible need for pursuing a new paradigm that moves beyond next-word prediction. We conclude with reflections on societal influences of the recent technological leap and future research directions."

0
0
0
https://community.openai.com/t/experiencing-decreased-performance-with-chatgpt-4/234269

For anyone else who wants to follow along with the community response to the recent quality degradation of ChatGPT.

0
0
0

"I apologize for the confusion." —ChatGPT

0
0
0

Confused the intro of Ascension by KOAN Sound with Autechre just now on a Spotify playlist, good job KOAN Sound! Nice sample selection, automation, and mastering.

0
0
0

I've noticed a significant quality degradation across my paid ML apps (e.g., Copilot, ChatGPT), which from what I've gathered, is due to each company's preemptive avoidance of potential legal complications related to their model's training data and copyright concerns—but the net result is that the perceived capability of Copilot, for example, has dropped significantly in the last couple of months.

There's a long delay between typing a line of text and the predictive hint appearing, and the hint itself is so watered down now that it's almost not useful anymore (especially when compared to workshopping code with ChatGPT-4, which has a similar monthly price point).

0
0
0

Coding really is a form of active digital hallucination. When I stare at the screen, my mind's eye can see the UI evolve, yet I'm also aware of the delay between my conceptualization and the actualization due to my fingers conjuring up the symbols required to realize it.

0
0
0

One of the biggest improvements to my #running practice was the decision to stop listening to music while I run. It's really helped me to be more mindful and to clear my head so that I can actually hear myself think. Also helps me to better understand my breathing patterns.

0
0
0

Profile update: Member profiles now include a list of their most frequently used tags.

0
0
0