Meet levels.fyi – they collect actual like for like data on salaries, benefits, levels etc. for the US tech industry.
They use this to help people negotiate salaries (how they monetise).
Levels recently published a report for 2021 that has some fascinating data (h/t The Diff).
The table attached shows entry level engineer salaries.
Lots of other interesting stats – comp has been rising (generally highest entry-level salaries are growing +3.4% annualised since 2019) and the Bay Area still wins (40% higher than LA for example).
Google has been running an internal prediction market – Gleangen, and Astral Codex Ten has a great write up on the topic.
This is the second iteration of such a market (the first was called Prophit).
Google claims that anyone can now build a prediction market on Google Cloud.
Prediction markets are fascinating as a tool but have struggled to get really big and more importantly to solve the three key issues (real money, easy to use, easy to create own markets).
Metaculus is a community dedicated to making accurate predictions (they have a great resource page) as is Manifold. Neither use real money.
Kalshi is a new startup ($30m of funding) that is trying to make events into an asset class via a real money prediction market. As is Futuur.
Polymarket, the biggest such market in the US, was recently fined and forced to shut down in the US (it remains open elsewhere).
You probably know, AlphaGo, the 2016 AI program that dominated the game of Go.
Soon after, a software implementation called Leela was made available, to train human Go players.
Data from 750k Go moves from 1,200+ players between 2015-2019 shows a significant improvement in move quality – especially among younger players (see chart) who are likely more open to learn from Leela.
Starship, the fully reusable rocket under development by SpaceX, is a revolution the industry grossly under-appreciates. So goes this fascinating blog post.
“Starship matters. It’s not just a really big rocket, like any other rocket on steroids. It’s a continuing and dedicated attempt to achieve the “Holy Grail” of rocketry, a fully and rapidly reusable orbital class rocket that can be mass manufactured. It is intended to enable a conveyor belt logistical capacity to Low Earth Orbit (LEO) comparable to the Berlin Airlift.“
“Consider the two critical metrics: Dollars per tonne ($/T) and tonnes per year (T/year) … Starship is intended to reach numbers as low as $1m/T and 1000 T/year for cargo soft landed on the Moon. Apollo achieved about $2b/T and 2 T/year for cargo soft landed on the Moon.“
It is developing in leaps – “Two years ago Starship was a design concept and a mock up. Today it’s a 95% complete prototype that will soon fly to space and may even make it back in one piece.“
Counter to the prevailing narrative – e-commerce hasn’t seen a step change and is almost exactly where a 10-year trend line would have predict it would be as a share of total retail in the US.
The reason for this is that total retail sales has grown strongly (+13% vs. normally being +2-3%). In absolute terms, shoppers spent $204bn on e-commerce in Q3 2021 but the pre-pandemic trend would have predicted $183bn.
Absolutely worth a look as it nicely covers macro and strategic trends from the tech industry.
This chart shows how software as a service (SaaS) has pushed up the amount of software companies can use (as there is no need to get IT to install/support/configure new applications).
The move to cloud delivery for software has miles to go (it is just 10-15% of enterprise IT spend and 20-30% of workloads).
COBOL is the programming language that underpins the entire financial system.
“Over 80% of in-person transactions at U.S. financial institutions use COBOL. Fully 95% of the time you swipe your bank card, there’s COBOL running somewhere in the background.“
“The second most valuable asset in the United States — after oil — is the 240 billion lines of COBOL”
The language is old (from the 1960s) and runs on huge machines (mainframes), yet it is extremely suited to the task of processing billions of transactions very fast.
“In conversations with POLITICO, more than a dozen industry and government officials involved with the work of Gaia-X said the project was struggling to get off the ground amid infighting between corporate members, disagreement over its overall aims and a bloated bureaucratic structure that is delaying decisions. One industry official closely involved in the work of Gaia-X called it a “mess.”“
