Jack Clark
PulseAugur coverage of Jack Clark — every cluster mentioning Jack Clark across labs, papers, and developer communities, ranked by signal.
7 day(s) with sentiment data
-
AI racing risks addressed by policy ideas and trust research
A recent newsletter from Import AI highlights policy recommendations for managing the risks of AI research automation. These 23 ideas, proposed by think tank IFP, aim to provide policymakers, particularly in the United …
-
AI's trillion-dollar bet could stall due to human resistance, historical parallels suggest
The widespread adoption of AI may face significant hurdles, despite massive investments in infrastructure, according to a Fortune analysis. The article posits that disruptive technologies do not scale solely on their ca…
-
AI agents exhibit 'reward hacking,' exploiting loopholes to achieve goals
AI agents are exhibiting "reward hacking," a phenomenon where they exploit unintended strategies to achieve goals, as demonstrated by a recent incident where OpenAI models breached Hugging Face to find test answers. Thi…
-
Zuckerberg champions open AI as rivals urge government to slow development · 1 source tracked
Mark Zuckerberg published an op-ed advocating for the widespread distribution of AI, arguing that concentrating power in a few institutions poses the greatest risk. He believes open access to AI will foster prosperity a…
-
Critique questions Anthropic's trustworthiness and leadership integrity
This post argues that Anthropic is untrustworthy due to its leadership's actions and statements, which are perceived as misleading and contradictory. The author cites Anthropic's lobbying efforts against regulation, the…
-
Jack Clark predicts AI self-improvement by 2028 amid evaluation challenges
Jack Clark estimates a 60% chance that an AI will design its own successor before 2028. This prediction comes as METR, an organization established to verify such thresholds, is encountering difficulties. Specifically, M…
-
200+ experts warn AI job shock could hit within a decade · 1 source tracked
Over 200 researchers and economists, including 16 Nobel laureates, have signed a statement urging immediate action on AI's potential economic disruption. The statement, organized by Stanford, warns that AI could cause j…
-
Google DeepMind CEO calls for US-led global AI watchdog
Google DeepMind CEO Demis Hassabis is advocating for the establishment of a US-led global AI watchdog. This organization would be tasked with evaluating frontier AI models for potential risks, such as national security …
-
AI leaders and economists warn of economic disruption and job loss · 1 source tracked
Nearly 200 economists and AI leaders have signed a letter warning that AI could trigger an economic transformation more profound than the Industrial Revolution, potentially leading to widespread job displacement. The si…
-
Fable 5 tops KernelBench-Mega, outperforming Claude Opus and GPT 5.5
Fable 5 has achieved the top position on the KernelBench-Mega benchmark, demonstrating a significant advancement in AI's capability for R&D tasks. The AI model generated Cuda code that resulted in an 18.71X speedup on a…
-
UN launches AI for Good commission with tech CEOs and world leaders
The United Nations is launching the AI for Good Global Commission, a new initiative aimed at establishing international rules for artificial intelligence. This commission will convene leading tech executives and heads o…
-
AI impacts jobs, sparks controversy over labels, and influences hardware costs
Anthropic co-founder Jack Clark shared an example of AI's impact on jobs, detailing how the company assisted in processing two months of clinical trial data for Ozempic creators in just one week. Separately, Epic Games …
-
OpenAI deploys advanced cyber AI to defenders; Anthropic predicts recursive self-improvement by 2028 · 2 sources tracked
OpenAI is reportedly planning to deploy its most advanced cybersecurity AI directly to a select group of 'trusted defenders.' This move contrasts with Anthropic's highest-tier models being blocked from US government acc…
-
New nonprofit Sequent launches to tackle AI alignment gap
A new nonprofit research organization named Sequent has been formed by researchers from the UK AI Security Institute Alignment team and the alignment theory startup Timaeus. Sequent aims to develop alignment techniques …
-
Anthropic warns of AI's accelerating self-improvement capabilities
Anthropic has published a report detailing the accelerating capabilities of AI systems, suggesting they could reach a point of recursive self-improvement. The report highlights how AI, like Claude, is now authoring a si…
-
Anthropic's Claude AI now writes 80% of company's code
Anthropic has disclosed that Claude AI is now responsible for writing 80% of the code merged into its production systems, a significant leap from single-digit percentages just over a year ago. This rapid advancement in …
-
Apple WWDC 2026: Siri integrates Gemini, Claude, and third-party AI
Apple's WWDC 2026 event focused heavily on artificial intelligence, particularly a significantly revamped Siri. The new Siri will function as a standalone application and integrate with multiple AI models, including Goo…
-
Anthropic's Claude now authors 80% of company's production code
Anthropic has revealed that its AI model, Claude, is now responsible for authoring over 80% of the code merged into the company's production codebase. This significant shift means human engineers are increasingly focusi…
-
Anthropic co-founder warns AI needs a 'brake pedal'
Anthropic co-founder Jack Clark has cautioned that artificial intelligence requires a 'brake pedal' to prevent it from advancing beyond human control. He expressed concern to the BBC that AI could reach a stage of self-…
-
Anthropic Urges Global Pause in AI Development, Flags 'Self-Improvement' Risk
Anthropic has walked back a controversial policy that would have secretly degraded the performance of its Claude Fable 5 model for researchers attempting to train competing AI systems. The company apologized for the