While GenAI has revolutionised production speed and cost, its impact on actual performance has remained a subject of intense debate. The new study, titled “AI Ads That Work: How AI Creative Stacks Up Against Humans,” analysed hundreds of thousands of live ads running on Realize, Taboola’s performance advertising platform, totalling more than 500 million impressions and 3 million clicks. — Read More
Recent Updates Page 18
Ads Candidate Generation using Behavioral Sequence Modeling
At Pinterest, ads are more than just advertisements; they are a vital part of the content ecosystem, designed to inspire users and connect them with products and ideas they love. Our goal is to surface the right ads at the right time, ensuring they seamlessly integrate into a user’s shopping journey and provide genuine value. To achieve this, understanding user behavior is paramount.
Delivering highly relevant ads in a dynamic environment like Pinterest presents unique challenges. Users’ interests and shopping intents evolve rapidly, making it crucial for our ad systems to adapt and anticipate their needs. Traditional ad targeting methods often rely on broad demographic data or static interest categories, which can fall short in capturing the nuanced and evolving nature of user behavior. — Read More
How I Structure My Data Pipelines: The Silver Layer
… Dimensional modeling is more important than ever.
The methodology has decades of literature behind it. The patterns are documented, the edge cases are known, and there’s no need to invent solutions from scratch. Facts and dimensions are composable primitives that mix and match to answer questions nobody has thought of yet. Paired with an ERD, tests, and naming conventions, Silver becomes something people can navigate without asking questions.
Gold models are the primary consumers of Silver. Every metric view, every wide table, every consumption artifact in Gold starts by referencing Silver facts and dimensions.
Overview
The Bronze Layer
The Silver Layer
China’s Military Uses Hawk and Wolf Behavior to Train AI Weapon Swarms
On January 23, China’s National University of Defense Technology demonstrated something that’s reshaping how autonomous weapons work: a single operator supervising over 200 drones simultaneously during urban combat exercises. The swarm operated with minimal human input, relying on what the People’s Liberation Army calls “effect-based control,” designed to function even when communication signals are jammed.
The technology didn’t emerge from traditional programming. It came from watching hawks hunt.
Engineers at Beihang University, a military-linked institution, observed how hawks select vulnerable prey and trained defensive drones to replicate that behaviour, according to The Wall Street Journal. In parallel tests, attack drones mimicked pigeons to evade threats. The result: in a five-versus-five combat simulation, the hawk-trained drones eliminated all opponents in 5.3 seconds, according to a patent filed in April 2024. — Read More
Apple’s AI Game is Misunderstood
Apple’s AI strategy has become a Rorschach test for the technology industry. Critics see a company falling dangerously behind. Needham analyst Laura Martin claims it is one to two years behind its competitors. But almost all of this commentary, whether bullish or bearish, focuses on the wrong question.
The standard narrative compares Apple’s AI capex to Microsoft’s, Apple’s Siri to Google’s Gemini, Apple’s foundation models to OpenAI’s GPT-4. By these metrics, Apple looks behind. But these comparisons assume Apple is trying to win the same race. The evidence suggests it isn’t. — Read More
Corollary Discharge Dysfunction to Inner Speech and its Relationship to Auditory Verbal Hallucinations in Patients with Schizophrenia Spectrum Disorders
Auditory-verbal hallucinations (AVH)—the experience of hearing voices in the absence of auditory stimulation—are a cardinal psychotic feature of schizophrenia-spectrum disorders. It has long been suggested that some AVH may reflect the misperception of inner speech as external voices due to a failure of corollary-discharge-related mechanisms. We aimed to test this hypothesis with an electrophysiological marker of inner speech.
… This study provides empirical support for the theory that AVH are related to abnormalities in the normative suppressive mechanisms associated with inner speech. This phenomenon of “inner speaking-induced suppression” may have utility as a biomarker for schizophrenia-spectrum disorders generally, and may index a tendency for AVH specifically at more extreme levels of abnormality. — Read More
The Personal AI Mentor Setup I Wish I Had at 20
On This Day… 1776
On This Day… 1776 is Darren Aronofsky’s short-form series focusing on key moments from that revolutionary year. The fact-based short films use a “combination of traditional filmmaking tools and emerging AI capabilities,” SAG voice actors and AI visuals, to dramatize scenes from the year’s most pivotal moments. The series draws on Aronofsky’s Primordial Soup’s partnership with Google DeepMind having each episode drop on the 250th anniversary of the event it depicts. — Read More
Conditional Memory via Scalable Lookup:A New Axis of Sparsity for Large Language Models
While Mixture-of-Experts (MoE) scales capacity via conditional computation, Transformers lack a native primitive for knowledge lookup, forcing them to inefficiently simulate retrieval through computation. To address this, we introduce conditional memory as a complementary sparsity axis, instantiated via Engram, a module that modernizes classic N-gram embedding for O(1) lookup. By formulating the Sparsity Allocation problem, we uncover a U-shaped scaling law that optimizes the trade-off between neural computation (MoE) and static memory (Engram). Guided by this law, we scale Engram to 27B parameters, achieving superior performance over a strictly iso-parameter and iso-FLOPs MoE baseline. Most notably, while the memory module is expected to aid knowledge retrieval (e.g., MMLU +3.4; CMMLU +4.0), we observe even larger gains in general reasoning (e.g., BBH +5.0; ARC-Challenge +3.7) and code/math domains~(HumanEval +3.0; MATH +2.4). Mechanistic analyses reveal that Engram relieves the backbone’s early layers from static reconstruction, effectively deepening the network for complex reasoning. Furthermore, by delegating local dependencies to lookups, it frees up attention capacity for global context, substantially boosting long-context retrieval (e.g., Multi-Query NIAH: 84.2 to 97.0). Finally, Engram establishes infrastructure-aware efficiency: its deterministic addressing enables runtime prefetching from host memory, incurring negligible overhead. We envision conditional memory as an indispensable modeling primitive for next-generation sparse models. — Read More
The 80% Problem in Agentic Coding
… Some time ago I wrote about “the 70% problem” – where AI coding took you to 70% completion, then leave the final 30% last mile for humans. That framing may now be evolving. The percentage may shift to 80% or higher for certain kinds of projects, but the nature of the problem changed more dramatically than the numbers suggest.
Armin Ronacher’s poll of 5,000 developers compliments this story: 44% now write less than 10% of their code manually. Another 26% are in the 10-50% range. We’ve crossed a threshold. But here’s what the triumphalist narrative misses: the problems didn’t disappear, they shifted. And some got worse. — Read More
Rick's Cafe AI 