VentureBeat

Obsessed with covering transformative technology.

[VentureBeat @VentureBeat]

Why PayPal added proprietary business context and kept it separate from the model that uses it.
#AI #GenAI #AIAgents #LLMs

https://t.co/LzIMYk0pyQ

[VentureBeat @VentureBeat]

Tomorrow after work in SF: Snowflake, Atlassian and WGU on what it takes for teams to keep pace with AI.

Then drinks with other AI leaders.

Only a few seats left. Reserve yours today: https://t.co/eBlHscbAf7

[VentureBeat @VentureBeat]

TrustedRouter launches three new encrypted 'fusion' models that combine multiple LLMs, including one exclusively with U.S. model providers
https://t.co/BwHK6xcdHY

[VentureBeat @VentureBeat]

OpenAI lets you claim your business in ChatGPT and set up a custom agentic AI customer sales rep
https://t.co/cj0cTcELJU

[VentureBeat @VentureBeat]

Model churn is expensive. PayPal’s answer: keep the business knowledge — golden queries, analytical playbooks — outside the LLM, and just plug in the next better model.
https://t.co/IVwjeshdzF

[VentureBeat @VentureBeat]

Two days out. One room of AI and IT decision-makers, plus leaders from Snowflake, Atlassian and WGU. After work this Thursday in SF.

Requests close Wednesday night: https://t.co/eBlHscbAf7

[VentureBeat @VentureBeat]

Google Gemini 4 Carbon is now publicly available — and it undercuts Claude Fable 5.2 and Opus 5.5 on API price
https://t.co/S74d0KzlJ8

[VentureBeat @VentureBeat]

T-Mobile discovered their AI unexpectedly hallucinating hold music during routing pauses. Learn why large-scale interactive voice deployments require strict prompts to prevent autonomous behavior.

For more insights attend our next VentureBeat event on Thursday November 12.

Space is limited. Request your invite: https://t.co/tCpjJI2e4w

[VentureBeat @VentureBeat]

Why more GPUs and tokens don’t guarantee better AI agent outcomes
https://t.co/FQRVnRKPkt

[VentureBeat @VentureBeat]

AI’s biggest bottleneck is cost: Companies blame stalled projects on overruns three times as often as on compute shortages
https://t.co/veQwfsVku8

[VentureBeat @VentureBeat]

Did Claude Code’s ‘don’t ask again’ option accidentally expose your enterprise's passwords and keys?
https://t.co/O4ZrtekrNH

[VentureBeat @VentureBeat]

Meta’s proprietary Muse Ember nearly matches Claude Haiku 5.5 at 1/3 the estimated per-task cost
https://t.co/ol5fk1s9D6

[VentureBeat @VentureBeat]

Meituan releases the open weights of its LongCat-2.5 LLM with vision under a permissive MIT License
https://t.co/OdKwrrtPqu

[VentureBeat @VentureBeat]

AI agents miss what people know. Of October respondents whose organizations run AI agents on company data, 78% traced a confident but wrong answer to missing or inconsistent business context in the past six months. (Source: VB Pulse - Agentic Context Layers survey, October 2026)

Keeping Pace With AI is an intimate, in-person evening for enterprise technology leaders, hosted by VentureBeat Editor-in-Chief Matt Marshall.

Space is limited. Request your invite: https://t.co/tCpjJI2e4w

[VentureBeat @VentureBeat]

Getting an open model running is one thing. Keeping it fast, cost-effective and maintained is another. These four questions help technical teams weigh deployment, capacity, serving-stack ownership and GPU requirements before going into production. @Azure @AMD
https://t.co/J2UNBafafJ

[VentureBeat @VentureBeat]

Anthropic releases Claude Fable 5.2 and Mythos 5.2, pairing new, record-high benchmarks with 20% API cost reduction
https://t.co/zD1HhbHh3G

[VentureBeat @VentureBeat]

With decision models, we make a conscious choice not to include verbose reasoning traces and get straight to the answer. This System 1 thinking is very useful in specific scenarios.
#AI #GenAI #GenerativeAI #AIAgents #LLMs

https://t.co/UTch4BdLRr

[VentureBeat @VentureBeat]

AI made code faster. Now the bottleneck is leadership.

@Snowflake’s Will Allen on moving an entire engineering org to AI-native work, and the processes that had to change.

For enterprise AI leaders. Thu Nov 12, SF. Final seats: https://t.co/eBlHscbAf7

[VentureBeat @VentureBeat]

How to run open models in production without owning the serving stack
https://t.co/AL4nS4Bxbc

[VentureBeat @VentureBeat]

Decision models like Jev don’t explain their answers. That makes context your problem
https://t.co/fhsRrDStjM