2 hours ago
Bitget CEO Gracy Chen says AI trading apps are similar interns for now; soon they volition beryllium afloat clip employees.
Feb 11, 2026, 3:26 p.m.
Today’s AI trading bots are based connected a constricted magnitude of humanities information which means wholly unfamiliar marketplace events similar the 10/10 liquidations oregon adjacent past week’s terrible selloffs volition permission agentic trading models abbreviated of the mark.
These humanities data-driven AI models person ne'er seen immense liquidations connected a azygous time and would find this “very unfamiliar” said Bitget CEO Gracy Chen connected a sheet astir agentic trading bots astatine Consensus Hong Kong 2026. As specified quality involution is needed.
“As an exchange, we don’t program to physique our ain LLM [large connection model]. But trading bots are a large thing,” Chen said. “Current AI bots are a spot similar an intern: faster, cheaper but needs a small supervision.”
However, further down the enactment this volition beryllium much similar a “full employee,” and successful 3-5 years AI tin regenerate a batch of us, Chen said.
These are sentiments heard regularly successful the algorithmic trading satellite erstwhile it comes to AI.
While analyzable LLM and instrumentality learning trading exertion is improving astatine a accelerated clip, determination are inactive plentifulness of radical who deliberation a quality overlay is an indispensable portion of the process – peculiarly successful situations similar the terrible volatility that precocious gripped crypto markets.
Joining Chen connected the panel, Saad Naj, laminitis and CEO of agentic trading startup PiP World agreed the tech is successful its infancy and that comes with risk. But helium pointed retired that 90% of time traders and retail players suffer money.
“As humans we are excessively emotional. We can’t vie with AI solutions,” Naj said.
More For You
Decentralized AI is successful a trough but existent opportunities are emerging, crypto VCs say
Moving distant from overhyped GPU marketplaces and ample AI exemplary alternatives towards purpose-built, full-stack solutions is the goal.
What to know:
- The crypto-AI intersection is entering a quieter, much selective signifier with a displacement towards utility-driven applications, crypto VCs said.
- Moving distant from overhyped GPU marketplaces and ample AI exemplary alternatives towards purpose-built, full-stack solutions is the goal.
- Founders request to connection much than conscionable a wrapper connected existing AI models similar ChatGPT.
English (US) 