Layoffs, Dev Containers, Docker and GenAI
News #4: In this episode: StackOverflow cuts 28% of their stuff, CodeSandbox now supports Dev Containers, Docker introduces GenAI Stack project
StackOverflow cuts 28% of staff
StackOverflow has just announced cutting 28% of staff.
I believe the primary reason for the layoff has to be attributed to the drastic drop in page views in the last year due to the use of ChapGPT and other GenerativeAI models, as I reported in a previous episode of this newsletter News #2.
The company went on a hiring spree in 2022 when it doubled the headcount. Like many other companies, it has to lay off part of the staff to account for the current global economic situation.
We have seen numerous layoffs in the last couple of years, but this one still feels different.
First of all 28% is quite a big percentage even if the total number of people impacted is much less than other layoffs.
Second, given the recent adoption from StackOverflow of Generative AI in a project called Overflow AI, you would expect they would double down on their efforts to make search easier on their website. Many companies, that were involved in layoffs at the end of last year, have started hiring again in an effort to ride the wave of Generative AI. Some of them can already see huge increases in their stock prices and a good return on investments.
Lastly, if StackOverflow traffic were to slowly disappear, the same GenerativeAI models that are making the website obsolete would see their source of training data disappear with it.
In an effort to prevent this situation, a few months ago StackOverflow asked AI company to pay for scraping their website.
Now it looks like StackOverflow's effort to bring back some of that traffic with their OverflowAI project was maybe too little too late. Only time will tell.
CodeSandbox now supports Dev Containers
CodeSandbox, the famous web development tool favoured by many front-end engineers, has just announced support for Dev Containers
First of all, what is CodeSandbox?
CodeSandbox is an online development environment that allows developers to quickly create, edit, and share web applications in various programming languages and frameworks within a web browser. It provides an integrated code editor, live preview, and collaboration features, making it a convenient tool for web development and collaboration.
We discussed Dev Containers in a previous episode of this newsletter, where I compared different ways to develop inside a container.
Here I wanted to point out how another famous project now supports this open-source specification first introduced by Microsoft and then adopted by Github Codespaces and DevPod.
Docker introduces GenAI Stack project
At the beginning of the month, during their annual DockerCon, Docker announced a new GenAI stack project that should make it easier to work with GenerativeAI and LLM (Large Language Models) locally on your laptop, thanks to a partnership with Neo4j, LangChain, and Ollama,
The technologies involved in this new project are:
LangChain, a famous Python framework that facilitates creating and orchestrating applications that use LLMs.
Neo4j, a famous graph database that can be used as a vector store for semantic search.
Ollama, a framework to manage and run LLMs locally.
Docker compose, a way to start and connect more than a single container without the complexity of Kubernetes.
I haven't used this new tool from Docker yet, but from what I can read I don't think it is particularly innovative. You can use those same technologies in isolation already, without being limited to working on your laptop, to Docker compose. or a particular choice of vector database.
Given how every single company in the market is trying to find interesting uses for GenAI, I see this news from Docker as an extra incentive to invest more time into learning about these technologies.
I have only recently started investing my time into LangChain and LLMs models to create applications.
Until now I have only used ChatGPT as a final user.
Personally, I think, the current wave of Generative AI is not different from what Big Data and data lakes were 5-10 years ago. Now we take those technologies for granted. So I'll personally be riding this wave like I have done with Big Data. 😄
Especially since Elastic has now invested in Generative AI with a few different initiatives.