Artificial intelligence is essential in all sectors, sometimes as an ally, sometimes as a silent tyrant. Can we still stop it? Or do we live the start of a real rush to digital gold? At Coinbase, the answer is already clear: accelerating, even if it means shaking up its teams. Exchange Crypto assumes having taken a decisive step by entrusting almost half of its code to AI, and its CEO Brian Armstrong pushes even further.
In short
- Brian Armstrong confirms that 40 % of the Coinbase code is generated by IA.
- The objective set is to exceed 50 % IA code by next October.
- Armstrong imposed the compulsory use of Copilot and Cursor under threat of dismissal.
- Crypto recruitments retreat while AI now attracts talent and global investments.
Coinbase transforms its code into an AI laboratory
At Coinbase, AI is no longer an experiment, but a daily reality. According to Brian Armstrong, 40 % of code lines are now generated by AI, against less than 20 % a few months ago. The objective is clear: Switch to 50 % by October. For Armstrong, ” We must use AI responsible for as much as possible ».
The CEO also highlights spectacular results. At Coinbase, some engineers are now able to refact, update or build new systems in a few days, when it was necessary for months. This shortening of deadlines illustrates the promise brandished by Armstrong: a multiplied productivity, capable of redrawing the way in which an crypto exchange built and maintains its infrastructure.
THE Tools used There are many: Github Copilot, Cursor, Claude Code, Cody, Jetbrains. Over 1,500 engineers use it daily. The Crypto Exchange even follows the consumption of tokens as a productivity indicator.
By displaying these figures, Coinbase sends a strong message to investors: the listed company wants to be perceived as a pioneer of automation in Crypto finance.
Armstrong, a boss who imposes forced march AI
Behind this transition hides a brutal method. Armstrong gave his engineers A week to adopt Github Copilot and Cursor.
I said to myself: AI is important. You must all learn it and at least initiate it. You don't need to use it every day right away, we will do training, but you have to adopt it by the end of the week at least … Some had a good reason, because they just returned from a trip or other. Others did not have, and they were dismissed.
These words shocked part of his teams, but Armstrong assumes this turn. For him, the future of Coinbase is based on a “a-native” workforce, formed to work with these new tools.
The Crypto Exchange has also set up “AI Speedruns”: each month, an employee who brilliantly integrated the AIA to convey a workshop to transmit his practices.
If this approach worries by its brutality, Armstrong presents it as a must. In a sector dominated by speed and competition, staying lagging behind the AI would be strategic suicide.
Crypto or “A-Native” engineers? The market changes
This mutation exceeds the exchange coinbase. Since 2022, Crypto recruitments have become scarce. As Raman Shalupau of Cryptojobslist explains, ” The developers and entrepreneurs go where there are money and excitement. At the moment, it is the AI that attracts both ».
Coinbase illustrates this rocking. Development teams are now working in a saturated AI assistant ecosystem, where speed takes precedence over traditional methods. But this race for efficiency raises concerns.
The massive use of AI has led to an increase in bugs. To answer them, Coinbase has created a “rest is sensitivity matrix” in order to limit the risks on sensitive projects.
Significant facts to remember
- 40 % of the Coinbase code is generated by AI, objective: 50 % in October;
- Over 1,500 engineers use Cursor and Copilot daily;
- Since 2022, Crypto recruitment has been back for the benefit of IA projects;
- Layers have sanctioned employees refractory to the IA adoption.
In this context, Coinbase is no longer just an Crypto exchange: it becomes a symbol of a new era where the engineer “ai-native” takes precedence over the classic crypto engineer. But remains a question: how much can this dependence on the machine code go without weakening security?
AI divides, in finance as elsewhere. On American campuses, it already arouses intense debates on learning and academic integrity. Meanwhile, Washington considers it an economic and strategic engine. Between fear and fascination, the border remains slim, and the future belongs to those who will learn the machine.
Maximize your Cointribne experience with our 'Read to Earn' program! For each article you read, earn points and access exclusive rewards. Sign up now and start accumulating advantages.
