721 place 0

762 New 'Markovian Thinking' technique unlocks a path to million-token AI reasoning

VentureBeat
VentureBeat · 10/20/2025 19:00 EDT

Researchers at Mila have proposed a new technique that makes large language models (LLMs) vastly more efficient when performing complex reasoning. Called Markovian Thinking, the approach allows LLMs to engage in lengthy reasoning without incurring the prohibitive computational costs that currently limit such tasks.The team’s implementation, an environment named Delethink, structures the reasoning chain into fixed-size chunks, breaking the scaling problem that plagues very long LLM responses. Initial estimat

To see detailed statistics for the news please log in »

Read the original

Add your comment
You must be logged in with Facebook to read and write comments.

A newsletter a day!

You may get 10 most important news around midday in daily newsletter. Press the button and we will send you the most important news only, no spam attached.

or register

LIKE us on Facebook so you won't miss the most important news of the day!

News from the same source
VentureBeat VentureBeat
Silicon Valley
George Avalos @ Silicon Valley 1 place · 02/07/2106 01:28 EDT

Newark apartment complex bought for much less than prior value

An East Bay apartment complex has been bought at a price that's well below its prior value. Read more

0

🔮
22.03.2026 ♉︎ Dear Taurus! Today promises to be bright and eventful, especially in the fields of love... Read more ›
Silicon Valley
George Avalos @ Silicon Valley 2 place · 02/07/2106 01:28 EDT

PG&E buys San Jose building to bolster South Bay operations

A PG&E Corp. unit has bought a San Jose building in a move to bolster the utility's South Bay operations. Read more

0

Habr
outlingo @ Habr 1 place · today 14:39 EDT

Им не поможет белый список

В школах начались каникулы, количество статей и обзоров решений по обходу блокировок и эксплуатации белых списков увеличилось вдвое. А мне уже надоели обсуждения механизмов обхода блокировок и шум в чатах и группах про блокировки, баны, переборы IP, белые списки и т.д. - и я решил обсудить не "как сделать чтобы", а "почему то о чем говорят не сработает".А есть и еще один вопрос - предположим, вы администратор какого-либо сервиса, внесенного... Read more

0 newcommer

Engadget
Jackson Chen @ Engadget 1 place · today 14:37 EDT

Crimson Desert developer apologizes and promises to replace AI-generated art

The developer behind the open-world RPG Crimson Desert has issued an official apology after players discovered several instances of AI-generated art in the game. Pearl Abyss posted on X that it released the game with some 2D visual props that were made with "experimental AI generative tools" and forgot to replace them before launch. We would like to address questions regarding the use of AI in Crimson Desert.During development, some... Read more

0 newcommer

Habr
EvgenyZinovyev @ Habr 2 place · today 14:19 EDT

ИТ в быту. Citroen C3 Picasso

Сегодня в ходе разговора с товарищем затронули тему о том, что нафига этот ваш ИТ вообще нужен. Решили поставить эксперимент: товарищ предлагает любую задачу, а я рассказываю нужны ли ИТ. Читать далее Read more

0 fresh

Habr
ANTON62 @ Habr 3 place · today 14:15 EDT

Когда клиентом становится AI: как проектировать инфраструктуру для A2A-коммерции

Весь современный веб спроектирован для одного типа клиента — человека с браузером. Но когда клиентом становится AI-агент, оказывается, что большинство привычных решений работают неправильно или не работают вообще. В этой статье — разбор архитектурных проблем, которые возникают при проектировании инфраструктуры для Agent-to-Agent (A2A) взаимодействия: адресация, маршрутизация, доверие и поиск. Читать далее Read more

0 fresh

Habr
SLY_G @ Habr · today 14:05 EDT

Астрономы создали одну из самых точных и всеобъемлющих космических карт, когда-либо составленных, раскрывшую блестящее «море света», которое пронизывало раннюю Вселенную.В отличие от других карт Вселенной, это трёхмерное изображение составлено из света, излучаемого одним-единственным элементом: водородом, самым простым и распространённым элементом во Вселенной. Он излучает большие количества света определённой длины волны, когда возбуждается энергией близлежащих звёзд.Измерив этот свет на обширном участке н Read more

0 fresh

Business Insider
Tristan Anthony,Lauren Edmonds @ Business Insider 1 place · today 14:04 EDT

Travelers describe airport chaos as unpaid TSA agents stop showing up: 'The line is coming from all directions'

Amid mounting disruptions, President Donald Trump on Saturday said he would send ICE agents to replace TSA officers at airports nationwide. Read more

0 fresh

CoinDesk
Nikhilesh De @ CoinDesk 1 place · today 14:00 EDT

The SEC explains how it's viewing a crypto security: State of Crypto

Joint SEC-CFTC interpretive guidance outlines how the agencies will determine whether a cryptocurrency is a security. Read more

0 fresh

Slashdot
EditorDavid @ Slashdot 1 place · today 13:55 EDT

William Shatner Celebrates 95th Birthday, Smokes Cigar, Revisits 'Rocket Man' and Tests X Money

It was 60 years ago when William Shatner — born in 1931 — portrayed Captain Kirk in the TV series Star Trek. Shatner turns 95 today — and celebrated by posting a picture of himself smoking a cigar. "At 95, I'm still smokin'!" Shatner joked, adding that in life he'd learned two things. "Never waste a good cigar. Never trust anyone who says you should 'act your age.'" For more... Read more

0 fresh

SlashGear
SlashGear 1 place · today 13:45 EDT

Why Does Harley-Davidson's Milwaukee-Eight Engine Have Four Spark Plugs?

The Harley-Davidson Milwaukee-Eight was designed as a successor to the company's V-twin bike engines, and it features unique engineering like four spark plugs. Read more

0 fresh

Habr
Creatman @ Habr · today 13:42 EDT

Как я собрал statusline для Claude Code с мониторингом VPS за одну сессию

Три VPS, MCP SSH, сессии по 10-15 часов. Каждый раз одно и то же: сколько контекста съедено? Пора /compact? Какой сервер под нагрузкой? Ни один из существующих statusline это не показывал. Собрал свой — за одну сессию в claude.ai.Вот как это выглядит в реальной рабочей сессии: Читать далее Read more

0 fresh

Habr
cognitronn (BotHub) @ Habr · today 13:34 EDT

[Перевод] Как китайские ИИ-модели обогнали американские по реальному потреблению

Призрачная модель под названием Hunter Alpha только что покорила мир ИИ.Никто не знает, кто её создал.Данные OpenRouter подтверждают сдвиг: китайские модели на прошлой неделе выдали 4,69 триллиона токенов. Американские — 3,29 триллиона. Вторую неделю подряд.Дочитайте до конца — и вы поймёте, почему эра агентов перевернула доску, и получите чек-лист, чтобы не строить на устаревшем стеке.Если вы выкатываете ИИ-фичи в продакшен, эти данные меняют вашу дорожную карту. Читать далее Read more

0 fresh

The Verge
Terrence O’Brien @ The Verge 1 place · today 13:34 EDT

Crimson Desert dev apologizes for use of AI art

Reviews of Crimson Desert have been mixed, but the bigger issue for the game has been the discovery of what appeared to be AI-generated assets in the final release. Now the developer has acknowledged that AI art was indeed used during the game's creation, but says that it was intended to be replaced before release. […] Read more

0 fresh

Eurogamer.net
Vikki Blake @ Eurogamer.net 2 place · today 13:31 EDT

New report claims Crimson Desert's development was plagued by "a hodgepodge of features crammed together" and leaders who "don't acknowledge anyone who doesn't share their exact mindset"

Two Pearl Abyss developers have claimed Crimson Desert was developed in a troubled environment, characterised by an inverted management hierarchy and leaders who "don't acknowledge anyone who doesn't share their exact mindset". Read more Read more

0 fresh

The most popular news from the same source for the last week
VentureBeat VentureBeat
VentureBeat
VentureBeat 3 place · 03/15/2026 15:00 EDT

Recent reports about AI project failure rates have raised uncomfortable questions for organizations investing heavily in AI. Much of the discussion has focused on technical factors like model accuracy and data quality, but after watching dozens of AI initiatives launch, I’ve noticed that the biggest opportunities for improvement are often cultural, not technical.Internal projects that struggle tend to share common issues. For example, engineering teams build models that product managers... Read more

0

VentureBeat
VentureBeat · 03/15/2026 18:00 EDT

Recent reports about AI project failure rates have raised uncomfortable questions for organizations investing heavily in AI. Much of the discussion has focused on technical factors like model accuracy and data quality, but after watching dozens of AI initiatives launch, I’ve noticed that the biggest opportunities for improvement are often cultural, not technical.Internal projects that struggle tend to share common issues. For example, engineering teams build models that product managers... Read more

0

VentureBeat
VentureBeat · 03/16/2026 00:00 EDT

LinkedIn's feed reaches more than 1.3 billion members — and the architecture behind it hadn't kept pace. The system had accumulated five separate retrieval pipelines, each with its own infrastructure and optimization logic, serving different slices of what users might want to see. Engineers at the company spent the last year tearing that apart and replacing it with a single LLM-based system. The result, LinkedIn says, is a feed that... Read more

0

VentureBeat
VentureBeat · 03/16/2026 01:00 EDT

Presented by AudioEyeWhile most organizations recognize the importance of accessibility from a theoretical angle, a stark gap exists between that awareness and actual execution. Companies can't just give a nod to accessibility -- and it can't just be a nice-to-have. The chasm between knowing and doing is not only exposing businesses to significant legal risk, it's also costing them actual business and growth opportunities. According to AudioEye’s newly released 2026... Read more

0

VentureBeat
VentureBeat · 03/16/2026 01:00 EDT

For more than two decades, digital businesses have relied on a simple assumption: When someone interacts with a website, that activity reflects a human making a conscious choice. Clicks are treated as signals of interest. Time on page is assumed to indicate engagement. Movement through a funnel is interpreted as intent. Entire growth strategies, marketing budgets, and product decisions have been built on this premise.Today, that assumption is quietly beginning... Read more

0

VentureBeat
VentureBeat · 03/16/2026 04:00 EDT

LinkedIn's feed reaches more than 1.3 billion members — and the architecture behind it hadn't kept pace. The system had accumulated five separate retrieval pipelines, each with its own infrastructure and optimization logic, serving different slices of what users might want to see. Engineers at the company spent the last year tearing that apart and replacing it with a single LLM-based system. The result, LinkedIn says, is a feed that... Read more

0

VentureBeat
VentureBeat · 03/16/2026 04:30 EDT

Jensen Huang walked onto the GTC stage Monday wearing his trademark leather jacket and carrying, as it turned out, the blueprints for a new kind of monopoly.The Nvidia CEO unveiled the Agent Toolkit, an open-source platform for building autonomous AI agents, and then rattled off the names of the companies that will use it: Adobe, Salesforce, SAP, ServiceNow, Siemens, CrowdStrike, Atlassian, Cadence, Synopsys, IQVIA, Palantir, Box, Cohesity, Dassault Systèmes, Red... Read more

0

VentureBeat
VentureBeat · 03/16/2026 12:00 EDT

An attacker embeds a single instruction inside a forwarded email. An OpenClaw agent summarizes that email as part of a normal task. The hidden instruction tells the agent to forward credentials to an external endpoint. The agent complies — through a sanctioned API call, using its own OAuth tokens. The firewall logs HTTP 200. EDR records a normal process. No signature fires. Nothing went wrong by any definition your security... Read more

0

VentureBeat
VentureBeat · 03/16/2026 15:14 EDT

Chinese AI startup Z.ai, known for its powerful, open source GLM family of large language models (LLMs), has introduced GLM-5-Turbo, a new, proprietary variant of its open source GLM-5 model aimed at agent-driven workflows, with the company positioning it as a faster model tuned for OpenClaw-style tasks such as tool use, long-chain execution and persistent automation. It's available now through Z.ai's application programming interface (API) on third-party provider OpenRouter with... Read more

0

VentureBeat
VentureBeat 3 place · 03/16/2026 16:30 EDT

Nvidia on Monday took the wraps off Vera Rubin, a sweeping new computing platform built from seven chips now in full production — and backed by an extraordinary lineup of customers that includes Anthropic, OpenAI, Meta and Mistral AI, along with every major cloud provider.The message to the AI industry, and to investors, was unmistakable: Nvidia is not slowing down. The Vera Rubin platform claims up to 10x more inference... Read more

0

Most popular sources

  • You see 372 news out of 372.
  • Sources 61 out of 61.
ScienceDaily 0%
The Next Web 0%
Vox 0%
Wired 0%
The Fintech Times 0%
View sources »

LIKE us on Facebook so you won't miss the most important news of the day!

22.03.2026 14:57
Last update: 14:50 EDT.
News rating updated: 20:50.

What is Times42?

Times42 brings you the most popular news from tech news portals in real-time chart.
Read about us in FAQ section.


Times42 © 2026