Kuvugururwa buri munsi1793 inkuru zagenzuwe
Source-checked AI coverage of product launches, policy shifts, safety research, and industry moves, explained in plain English by a nonprofit education team.
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
What happened, why it matters, and what to watch — without the jargon.
Iyo ikimenyetso ari gito, ntacyo dusohora aho gukanda ibiryo.
Source-checked AI stories, newest first, for people who need to understand AI without chasing hype.
AWS ivuga ko uburyo bwagutse bwo kwandika bwanditse bushobora guhindura amategeko yanditse muri politiki ya Dogwood igabanya ibikorwa byabakozi mugihe, kubahiriza ibisabwa nimbibi, kugenzura inyandiko-yubusa hamwe na Bedrock Guardrails, no kubuza ibikoresho byinjira.
Microsoft Ubushakashatsi buvuga ko Skala 1.1 itezimbere ubunyangamugayo kubayibanjirije mugihe igumye hafi yo kubara uburyo bwo kubara igice cyaho cya density-imikorere yuburyo bukoreshwa, none kikaba kiboneka muri CP2K hamwe nibiteganijwe kubindi bikoresho bya chimie nini.
microsoft.comInyandiko ya arXiv ivuga ko abakozi ba code bashobora gukora muburyo butandukanye kuri codebase ihwanye, hamwe ningaruka zitandukanye nurugero, urwego rwabakozi hamwe nigipimo.
Ubu bushakashatsi bwerekana icyitegererezo cyongeweho gihuza Fuzzy Logic nuburyo bwa IEEE Key Gas Method (FL-KGM) itangiza imikorere yabanyamuryango inonosoye, amategeko agenga fuzzy, hamwe no gutandukanya udushya twa CO na CO2 kugirango bikureho ibitagenda neza.
Ubushakashatsi bwa arXiv busuzuma imiterere yururimi rusanzwe kandi rusanzwe ku bipimo bitatu byo guhamagarira ibikoresho, bitanga ibisubizo bikomeye kumurimo wakazi bisaba guhamagarwa kwa API guterwa nuburyo bushoboka bwo guhuza n'imihindagurikire y'ikirere.
Urupapuro rushya rwa arXiv rutanga protocole-abakozi batatu-basubiramo protocole aho usubiramo asuzuma kode yabakozi hamwe nubugenzuzi bugenzura busubiramo mbere yo guhindura. Abanditsi batanze raporo y'ibisubizo byagezweho hejuru y'ibipimo byageragejwe, mu gihe banagaragaza ubwumvikane buke nk'uburyo bwo gutsindwa.
Inyandiko ya arXiv igereranya zeru-shoti na parameter-ikora neza-neza neza kugirango hamenyekane urwango-mvugo mu rurimi rwikiromani. Abanditsi bavuga ko kurwanya imihindagurikire y'ikirere byazamuye imikorere ya F1 kuva 0.56 ikagera kuri 0.93 kuri corpus irimo ibitekerezo bisaga 72.000.
Urupapuro rwa arXiv rutangiza FraudBench, igipimo cyo gusuzuma niba abakozi bakoresha amabanki bakoresha ibikoresho bashobora gutahura uburiganya bugaragara mubiganiro. Mu isuzuma ryibanze rimwe, abakozi bane batsinze 49% kugeza kuri 65% kumutekano wibitero.
A study of more than 2,000 language-model training runs says scarce target data can be repeated 15–20 times in mixtures, with the best rate varying by scale and compute.
Apple researchers describe LINK, a pretraining intervention that replaces selected English words with word-level translations from a target language. The paper reports improvements across eight languages and five model sizes, including up to a twofold speedup in reaching equivalent downstream performance.
A position paper reports tacit collusion by DeepSeek-R1 agents in a simulated Bertrand pricing market, even after human prompts against collusion. It argues that observed-behavior certification should precede deployment of reasoning agents in economic markets; the evidence and safeguards remain preliminary.
A systematic review surveys how large language models are being studied for mental-health analysis, risk assessment, therapy support and multimodal monitoring, while stressing unresolved ethical and regulatory challenges.
One useful briefing each week
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Kohereza akazi ka AI Tanga igikoresho cya AI