10 probability concepts for machine learning
a plain-language guide to ten essential probability ideas that make machine learning work, from random variables to entropy.
author
baris
Baris operates SummarizedData and maintains its sources, publishing rules, and automation. Read the editorial process.
savi app uses ai to block real-time phone scams
savi security launches a consumer app that detects ai-powered scams in texts, emails, and live calls, backed by $7 million in seed funding.
cora: per-slice orthogonal rotation for svd-based low-rank adaptation
cora introduces per-slice coherent orthogonal rotations to preserve pretrained geometry in svd-based fine-tuning, improving parameter efficiency.
us autonomous atvs deployed in ukraine combat
forterra has sent over 100 self-driving atvs to ukraine, marking the largest use of american autonomous ground vehicles in war.
benchmark audits can silently fail in five ways
a study shows that perturbation-based construct-validity audits can produce misleading conclusions due to hidden pipeline failures, and proposes a due-diligence gate to catch them.
graft adds per-word pronunciation control to text-to-speech
a new method lets text-to-speech models pronounce tricky words correctly by using short audio hints, without changing the speaker's voice.
automated data readiness for scientific ai
a new open-source framework called redi automates the transformation of raw scientific datasets into ai-ready formats with built-in reproducibility and scaling.
when pairwise comparisons fail for pluralistic preferences
local pairwise comparisons can misrepresent preferences when people hold multiple global priorities like fairness or proportionality.
siri gets pace and expressivity controls in ios 27 beta
apple's latest ios 27 developer beta lets testers adjust siri's speaking speed and emotional tone, part of a generative ai overhaul.
sk hynix ipo gives us investors a piece of the ai memory boom
south korean memory chipmaker sk hynix plans a us ipo, offering american depositary receipts to tap into ai-driven demand for high-bandwidth memory.
ai ransomware still needs a human operator
sysdig's jadepuffer attack used an ai agent for technical execution but a human still set up infrastructure and chose the victim.
pytorch monarch comes to amd gpus for fault-tolerant training
pytorch monarch now supports amd instinct gpus with rocm, enabling single-controller distributed training that recovers from node failures without full restarts.
tech layoffs in 2026 where ai was cited
a running list of major tech layoffs in 2026 where companies named ai as a factor, with details on cuts at microsoft, oracle, gitlab, google, and more.
vercel ceo on splitting models from agents
vercel ceo guillermo rauch discusses the shift from prototyping to production ai, the rise of coding and internal agents, and the need to decouple models from agent platforms.
google now trains ai on your media unless you opt out
a recent privacy update lets google save images, audio, and video from search services for ai training by default, but users can disable it.
microsoft cuts 4,800 jobs in xbox and sales
microsoft laid off nearly 5,000 employees, with xbox losing 1,600 roles, as the company restructures amid ai-driven changes and a gaming hardware crisis.
how photoroom built the data pipeline for prx
photoroom details its data strategy for the prx text-to-image model, covering dataset assembly, captioning, and storage formats.
data scientists shift to ai oversight roles
data scientists now spend more time managing ai systems than building models, with new roles in governance, prompt engineering, and agent supervision.