AI Judge changed title after evaluation, original title: "Ken Griffin, Citadel founder, says AI agents now complete complex market analysis and deal structuring tasks in hours or days that previously required teams of master’s and PhD holders working weeks or months"
Griffin spoke at the Stanford Leadership Institute.
if you think the work is done - walk done the street and have some imagination
Not to be “too” optimistic, but jobs aren’t going away because “the” job is never done. The world can always be better: more food to grow, disease to conquer, goods to create, and experiences to have.
AI to 10x cost savings <<< AI to 10x our ambition of what our world can be
Positive users celebrate AI agents automating PhD-level finance jobs at Citadel for huge productivity gains while negative users hate the technology and dismiss the claims as lacking evidence.
No Digg Deeper questions have been answered for this story yet.
Where will AI be in 1, 2 or 3 years?
Ken Griffin doesn’t understand the ceiling just got raised. Some 20-something maybe reading this will build the cracked AI-human-computer-symbiosis team that will supersede his whole operation because he is too distracted about lowering cost
Boil the ocean don’t cut your costs
Ken Griffin went home on a Friday "fairly depressed" after watching AI agents at Citadel do work that used to take teams of PhDs in finance months to complete. Done in days.
His words: "These are not mid-tier white collar jobs. These are extraordinarily high skilled jobs being automated by agentic AI."
This is the head of one of the most successful hedge funds in history saying the people he pays seven figures to analyze markets and structure deals are being replaced by software that works in hours instead of months. Not theoretically. In his own office. Right now.
The Coatue deck we covered earlier this week called agents "the biggest unlock" in AI. Griffin just confirmed it from the buy side. The shift from copilots to agents is not a future event. It is already happening at the highest levels of finance.
Full interview here:
https://youtu.be/Csjy_A3Kj9s?si=5hFZlp_L-nAs5aN3
People are conflating a complexity frontier shift with work being "solved".
There will always be a competitive frontier, especially in finance, defense, science, law, etc.
There will be jobs and competition to be the best AI-human hybrid org at the frontiers.
Ken Griffin doesn’t understand the ceiling just got raised. Some 20-something maybe reading this will build the cracked AI-human-computer-symbiosis team that will supersede his whole operation because he is too distracted about lowering cost
Boil the ocean don’t cut your costs
Startups using AI agents to arbitrage intelligence can compete with the world’s leading companies who aren’t leveraging AI agents fully yet.
This opportunity is fleeting, act quickly.
Ken Griffin doesn’t understand the ceiling just got raised. Some 20-something maybe reading this will build the cracked AI-human-computer-symbiosis team that will supersede his whole operation because he is too distracted about lowering cost
Boil the ocean don’t cut your costs
@DrBeavisAI Agentic AI
@TFTC21 Yeah he watched them do it but did he check the quality of the slop?
https://www.fundamentedge.com/ai-accelerator
I use AI daily as a series of analysts. It’s like having my own team on standby. However you need a fairly deep understanding of the fields and subfields to be sure you’re not getting hallucinatory agreeable slop.
This is gonna make us exceptionally vulnerable to rationalized versus reasoned answers, due to the loss of rigor in our educational system. I know many graduates from about 10-15 years after me who cannot determine falsehood from truth.
@FundamentEdge Wow he used to be anti AI calling it all hype. What changed ???
@FundamentEdge The ultimate irony of the knowledge economy is that the most expensive, highly specialized cognitive labor was actually the easiest to map out and automate.
@FundamentEdge We knew he would turn. Everyone will…just before everything makes a mid term top.
@FundamentEdge Claude code only really launched 1 year ago (sonnet 4.0). So we are now at the iPhone 3S stage. Imagine where we will be in 10 years.
In three years AI is the default execution layer for digital work.
One year from now, the change is obvious inside work but still uneven across society. Software engineering, research, writing, finance, marketing, support, recruiting, legal drafting, operations, analytics, and internal tooling are all agent-assisted by default in serious companies. The average worker still has a job title that sounds familiar, but the task stack underneath it has already changed. The people who use AI well look superhuman. The people who do not use it look slow.
The most visible labor shock in year one is hiring compression, not mass unemployment. Fewer junior roles. Fewer backfills. Smaller teams. More “productivity” language. More companies discovering that one strong operator plus agents can replace three to ten mediocre workers. Management will pretend this is augmentation while quietly changing headcount plans.
Two years from now, agentic workflows become normal inside elite firms. Not just chatbots. Actual work loops: pull data, write code, test, summarize, compare, create reports, update dashboards, draft contracts, reconcile accounts, screen candidates, monitor markets, generate customer responses, build internal apps, and escalate exceptions. The human becomes reviewer, commander, taste layer, and accountability node.
That is when the white-collar ladder starts visibly breaking. The entry-level layer gets hollowed first because agents can do the grind that used to train juniors. The senior people become more leveraged, but the pipeline that creates future senior people becomes unstable. This is the part institutions are not ready for.
Three years from now, AI is no longer treated as “software.” It becomes infrastructure. Companies will be judged by whether they have an agent operating layer, proprietary data context, workflow control, model routing, evaluation systems, security permissions, and human override. The winners will not be the companies with the flashiest demo. The winners will be the ones that turn AI into governed production.
The app layer gets slaughtered. Lightweight SaaS, generic tools, simple workflow products, and seat-based software without data gravity or control-plane status keep getting repriced. The middle of software takes the most damage. The top survives by becoming the rail through which agents act. The bottom gets generated on demand.
The model layer becomes more brutal too. A few frontier players remain, but pure model advantage gets harder to monetize cleanly because enterprises need deployment, governance, security, integration, and trust. The labs move deeper into services and workflow because raw intelligence alone does not capture the whole value chain.
Power and capital become the physical constraint. Data centers keep spreading. Electricity politics intensify. The public backlash grows because the social pitch starts sounding insane: higher power bills, higher tech capex, more layoffs, more billionaire wealth, and a promise that society will be better later. That narrative will not hold without visible benefits for ordinary people.
The biggest surprise: the economy can look strong while the labor market feels worse. GDP can rise because machine cognition increases output. Workers can still feel poorer, less secure, and more replaceable because the surplus flows to owners, platforms, infrastructure, and the people who command the systems.
By year three, the divide is clear.
People who can command AI become dramatically more valuable.
People whose work was mostly repeatable cognitive production get repriced.
Companies that own workflow, data, distribution, compute, power, and trust compound.
Companies selling generic software or generic labor get compressed.
The deepest truth: AI will not feel like one product category anymore.
It will feel like a new economic atmosphere.
Everyone will breathe it. Not everyone will own the oxygen.
@FundamentEdge Citadel can do this because their data was already structured and audited before agents entered the picture. Most enterprises are trying to do the same thing on top of 10 years of SharePoint and PDFs nobody can find. The capability gap is real. The precondition gap is bigger.
@FundamentEdge I recall being very surprised Griffin called AI for research hype last October. In my opinion there hasn't been a step change improvement in agentic capability since then; it seems like his perspective has just caught up to what was consensus almost a year ago.
@elonmusk prolly in the Redroom 🟥
@FundamentEdge It's guaranteed that he's deceived.
Plug in a reality-based definition of LLMs (e.g., including their stochastic nature) into the narrative and it collapses in absurdity.
@elonmusk Geoffrey Hinton had a vision before the world could even imagine it.
@FundamentEdge this is the shift people are underestimating
agents don’t just make one person faster
they change how work should be assigned, checked, routed, and owned
the management problem moves from “who can do this analysis”
to “what workflow can safely absorb this much leverage”