The Pulse #112: Similarities between AI bots using a computer and end-to-end testing
Also: Automated reasoning proves system correctness at AWS, Winamp code shows why software licenses are important, and more
The Pulse is a series covering insights, patterns, and trends within Big Tech and startups. Notice an interesting event or trend? Send me a message.
Today, we cover:
Industry pulse. Year-on-year growth of software engineering indicates there’s a “Great Stay”, “agents” is the new AI buzzword, Stripe buys stablecoin company for $1.1B, Apple’s Vision Pro struggles, and more.
Similarities between AI bots using a computer and end-to-end testing. Anthropic has released an AI bot that operates a computer like a human. Looking closer, this technology’s limits seem the same as those of end-to-end testing. In fact, some end-to-end testing startups using AI for this task could be ahead of Anthropic.
Automated reasoning at AWS. Automated reasoning is a formal method to validate the correctness of systems using mathematical proofs. AWS is using it a lot, including to harden its S3 file storage service. LLMs occupy most attention in innovative software engineering approaches, and automated reasoning seems like a very significant field. Unlike non-deterministic LLMs, automated reasoning is fully deterministic!
Winamp code shows why software licenses are important. Winamp released its source code on GitHub. In a cheeky move, its team added an absurd clause that means nobody can use those additions. Developers then discovered that Winamp appeared to be breaking GPL licenses for open source code and ignoring legal requirements.
1. Industry pulse
Software engineering’s year-on-year growth
In Tuesday’s article about the state of the software engineering market, it mentions 2023 was the only year since 2000 when the number of employed software engineers dropped, according to data from Live Data Technologies. Here’s an alternative visualization, which shows the growth percentage in the group tracked:
As context, Live Data Technologies tracks 775,000 software engineer profiles. Every year, their data set has more files added.