Shortwave — rajhesh.panchanadhan@gmail.com [Gmail alternative]

The study leveraged large language models (GPT-3.5) to identify and analyze critical aspects of emerging corporate risks using earnings call transcripts — with respect to Political, Climate-related, and AI risks.

More than 69,000 earnings call transcripts were used in the study, and the GPT model was instructed to make human-readable risk summaries ... See more

How can we make interacting with conversational models feel more natural?

Every conversational interface to a language model adopts the same pattern:

A chat history sidebar, with each conversation lasting just a few turns

New sessions always begin in a brand-new thread

Every user query must always elicit exactly one response

None of these assumptions ar... See more

The idea is: learn to prompt chatbots very well = get way better outputs.

Right now, intricate prompting is helpful for some tasks. But over time, we think it’s an overrated skill. Here’s why:

1. As AI models improve, they require less “engineered” prompts. DALL-E 3 is a great example of this (you get top-tier images with < 10-word prompts).

In this c... See more

When the CPython process starts up, it performs runtime initialization. The runtime contains critical data required for the operation of the rest of the virtual machine.

Next, the Python code gets parsed and compiled into bytecode instructions, this is called tier-1 bytecode. This compiled bytecode is interpreted by the tier-1 bytecode interpreter.

T... See more

Denormalization

Another way Reddit minimizes joins is by using denormalization.

They took all the metadata fields required for displaying an image post and put them together into a single JSONB field. Instead of fetching different fields and combining them, they can just fetch that single JSONB field.

This made it much more efficient to fetch all the ... See more

Jazz.ToolsPowerSyncFireproofAutomergeDXOSElectricSQLand of course, Berlin's own: Yjs .

0.1% Ideas: Turning Life Into A Video Game

​

​What idea do you think is true AND underpriced?

​

Here’s mine: The most productive people design life like a video game.

I had this red pill moment when I saw the following:

The laziest person I’ve ever met play a video game for 16 hours straight, 7 days per week.

My brain began to hurt: The laziest person I k... See more

Lateral reading is a strategy that enables people to emulate how professional fact checkers establish the credibility of online information. It involves opening up new browser tabs to search for information about the organisation or individual behind a site before diving into its contents. Only after consulting the open web do skilled searchers gau... See more

Could you have GPT-4 ask questions to the user rather than force the user to do prompt engineering? An experiment says this gave superior results. My read is that they tested this on easy cases where we already knew it would work, so we did not learn much.