TechTalks

TechTalks

Home
Archive
About

July 2025

LegalPwn: The prompt injection attack hiding in the fine print of your code
LegalPwn, a new prompt injection attack, uses fake legal disclaimers to trick major LLMs into approving and executing malicious code.
 • 
Ben Dickson
1
MIT introduces new RL technique that moves beyond binary rewards
AI models are often overconfident. A new MIT training method teaches them self-doubt, improving reliability and making them more trustworthy.
 • 
Ben Dickson
1
How Windsurf came back from the brink
Back from the brink, Windsurf is "friends again" with Anthropic, integrating Claude Sonnet 4 and Devin in a bold new survival strategy.
 • 
Ben Dickson
New research reveals critical flaw in LLM-as-a-judge methods
Researchers discover critical vulnerability in LLM-as-a-judge reward models that could compromise the integrity and reliability of your AI training…
 • 
Ben Dickson
A first look at the new ChatGPT agent (and how it can change the internet)
OpenAI's powerful new ChatGPT Agent redefines AI capabilities while introducing new risk and attack vectors in security and data integrity of AI…
 • 
Ben Dickson
2
Inside the semantic attack that fools Grok-4 (and other LLMs)
Researchers jailbroke Grok-4 using a combined attack. The method manipulates conversational context, revealing a new class of semantic vulnerabilities.
 • 
Ben Dickson
1
A reality check on the 'emergent abilities' of LLMs
A new paper argues that "emergent abilities" in LLMs aren't true intelligence. The difference is crucial and has implications for real-world…
 • 
Ben Dickson
2
How Windsurf became a casualty of the AI arms race
Caught between tech giants, AI startup Windsurf became the main casualty when Google poached its leaders, gutting OpenAI's $3B acquisition deal.
 • 
Ben Dickson
1
Tackling the challenges of AI in space
An AI engineer explains how a new breed of tiny, hyper-efficient ML models is revolutionizing space, extending satellite life and unlocking the next…
 • 
Ben Dickson
Does aligning LLMs with human cognition come at the cost of less powerful models?
To make AI more human-like, must we sacrifice its power? A new study shows why LLM efficiency creates a gap in understanding.
 • 
Ben Dickson
1
© 2025 Ben Dickson
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture