The Threshold ReportAI progress, only when it's real.
Saturday, October 31Real: 9 · Security: 3 · Other News: 3 · Upcoming: 3
The Threshold Report
← Threshold Report

What’s Real

Robots+5Oct 31

Image: The Threshold Report/GPT Image 2.5

TU Delft's racing drone beats three human world champions

TU Delft's MonoRace won the AI Grand Challenge at the April 2025 A2RL Drone Racing Championship, according to IEEE Spectrum. The autonomous drone used one onboard camera and an inertial sensor (which measures motion), with an AI model turning estimates of its flight state into motor commands. It then beat three human world champions in first-person-view drone racing (where pilots steer through a live camera feed).

SOURCES · spectrum.ieee.org
Science+5Oct 31

Image: The Threshold Report/GPT Image 2.5

OpenAI opens GPT-Rosalind-6 to approved life-sciences researchers

OpenAI deployed GPT-Rosalind-6 to vetted researchers today for biology and drug-discovery work under access controls and monitoring. It uses GPT-6 Astra's learned model settings and refuses fewer biology questions that could have beneficial or harmful uses. OpenAI reports 49.6% versus Rosalind-5.5's 28.9% on its medicinal chemistry benchmark, and 37.7% versus 16.7% on its quantitative biology benchmark. OpenAI classifies it as high capability for biological and chemical risks and critical for cybersecurity, and says safety testing through simulated attacks is ongoing.

SOURCES · deploymentsafety.openai.com
Agents+2Text+2Oct 30

Image: The Threshold Report/GPT Image 2.5

Microsoft releases Mage-VL-Cowork for clicks and code

Microsoft released Mage-VL-Cowork yesterday for computer tasks that call for clicking through an interface or using code. Developers can download and adapt the compact model's 4 billion learned settings. Microsoft reports a 66.0% score on the OSWorld-Verified benchmark.

SOURCES · huggingface.co
Agents+2Science+2Oct 31

Image: blogs.nvidia.com

AI agents steer telescope observations from everyday-language requests

A request in everyday language can direct AI agents at the Allen Telescope Array to aim antennas, adjust measurements for accuracy and analyze incoming data. Live processing helps researchers change observations and search large data streams for interesting signals. In the Crab Nebula, the system detected a pulsar (a spinning remnant of a star) in what NVIDIA calls the first real-time AI detection of one.

SOURCES · blogs.nvidia.com
Agents+2Oct 29

Image: The Threshold Report/GPT Image 2.5

Zest launches restaurant recommendations and bookings inside text conversations

Zest launched a restaurant concierge that makes recommendations and books tables inside individual or group texts. It collects votes and reserves tables through services including OpenTable and Resy. The agent personalizes suggestions by using Plaid to import customers' food-and-drink transaction histories. TechCrunch found its recommendations worked fairly well in testing.

SOURCES · techcrunch.com
Audio+2Video+2Oct 30

Image: The Threshold Report/GPT Image 2.5

NVIDIA engineer builds a talking Halloween house with local AI

An NVIDIA engineer built an interactive Halloween house that responds with a customized voice character using models running on a workstation. The system combines speech recognition and voice synthesis with a language model and facial animation. A local agent and a larger language model created and judged more than 5,000 example conversations to train character adapters (model customizations for the characters). The builder tested models and adapters for quality, repetition and speed, reporting a median 1.70 seconds between the end of speech and generated audio.

SOURCES · huggingface.co
Compute & Infra+2Oct 31

Image: theverge.com

Google runs Gemma 4 aboard its Suncatcher satellite

Google's Suncatcher satellite ran Gemma 4 in orbit and returned a generated haiku today in the first operational test of its proposed space-computing system. The prototype also ran other tests, including prompts involving spacecraft measurements. The satellite launched in October 2026 carrying four TPUs (Google's processors for AI workloads).

SOURCES · techcrunch.com · theverge.com
Compute & Infra+2Oct 30

Image: cloud.google.com

Google Cloud releases automatic spending caps

Teams using Google Cloud can automatically pause eligible pay-as-you-go services when a project reaches its budget, with spend caps made generally available yesterday. The caps include an API, live estimated-cost reports and automatic monthly resets. Google says caps for AI services trigger within minutes of usage.

SOURCES · cloud.google.com
Robots+2Oct 29

Image: labs.scale.com

Scale AI and Hugging Face release step-labeled robot demonstrations

On Oct. 29, Scale AI and Hugging Face released FineART, letting robot developers train with instructions for individual steps within a task. The dataset contains 40,543 recorded episodes across 151 tasks, with 533,913 timestamped labels marking those steps. The partners also released FineART-VLA, a robot model that predicts task steps before translating them into actions. Scale AI says it evaluated the model in 3,400 real-world trials.

SOURCES · labs.scale.com · huggingface.co · huggingface.co
1 of 9

Curious whether AI can do something yet? Submit a question and we’ll track it for you.

See the board

Security

Money Moves

Market Watch

The AI stocks Wall Street fund managers are adding, trimming, or dropping.

Open the tracker for the latest holdings.

Open Market Watch
AI Entanglements

Who invests in, buys, supplies and powers whom across AI.

Open AI Entanglements
AgentsOct 29

Image: cursor.com

Tampered Claude Code extensions reach Cursor's marketplace

Cursor says tampered Claude Code extension versions 2.1.312 and 2.1.313 were served on Oct. 29 potentially affecting up to 5,000 users. Microsoft found attempts to establish lasting remote access and send out machine details, including the names of software environment settings without their values. Cursor removed the packages and blocked their server, Anthropic released version 2.1.314, and Microsoft says Anthropic told it that Anthropic's systems were not compromised. Developers who installed either affected version need to investigate their computers as well as update the extension.

Negative - Tampered Claude Code extensions were served through Cursor's marketplace, letting attackers distribute malicious code through a trusted software channel.
Positive - Cursor removed the malicious packages and blocked their server, cutting off the attack channel while Anthropic published a replacement extension.
SOURCES · cursor.com · microsoft.com · arstechnica.com
AgentsOct 29

Image: The Threshold Report/GPT Image 2.5

US charges five in alleged AI-assisted student-aid fraud schemes

Federal prosecutors charged five people in alleged student-aid fraud schemes involving AI on Oct. 29. Officials say participants used AI to complete applications and coursework and simulate attendance, maintaining fraudulent student identities over multiple semesters. The Education Department's inspector general first detected suspicious patterns, leading to an FBI investigation. Officials identified more than $32 million in actual losses tied to the subjects and seized more than $750,000 in cryptocurrency associated with a Kenyan network.

Negative - Fraudsters used AI to complete applications and coursework and fake attendance in student-aid schemes tied to more than $32 million in actual losses.
SOURCES · arstechnica.com
Compute & InfraOct 30

Image: blog.cloudflare.com

Cloudflare uses AI to locate systems exposed to weak keys

Within four hours of learning about weak eMudhra certificate keys on October 19, 2026, Cloudflare used AI-assisted source-code analysis to identify affected systems. The keys can be used to forge digital certificates (proofs computers use to verify a connection's identity), and Cloudflare observed forgeries within an hour of public disclosure on October 20, 2026. Cloudflare has patched all high-priority systems, though a small number of low-impact fixes remain. The company says it found no evidence that its own connections were intercepted.

Positive - AI-assisted analysis helped Cloudflare identify affected systems within four hours and patch all high-priority exposures from weak certificate keys.
SOURCES · blog.cloudflare.com
1 of 3

Other News

AgentsOct 29

Image: The Threshold Report/GPT Image 2.5

DoorDash discloses hundreds of orders placed by outside AI agents

Outside AI agents have placed several hundred DoorDash orders in total, CEO Tony Xu said on Oct. 29. His figure covers orders since those agents became available. DoorDash handled just under one billion orders overall in the third quarter of 2026 alone.

SOURCES · arstechnica.com
Compute & InfraOct 31

Image: The Threshold Report/GPT Image 2.5

Anthropic adopts shared website preferences for AI training

Website publishers can express AI-training preferences to multiple providers through a proposed robots.txt setting that Anthropic has adopted. The setting lives in robots.txt (a website file with instructions for automated visitors), reducing the need for separate preferences for each participating AI crawler. OpenAI plans to implement the signal in the months after October 2026, while Google and Microsoft plan support in 2027.

SOURCES · openai.com · theverge.com
Compute & InfraOct 31

Image: cloud.google.com

AI21 cuts training waits and manual scheduling on Google Cloud

AI21 reduced the time its high-priority Jamba training jobs waited for computing resources from 72 hours to 12, according to Google. The company moved the work to Google Cloud's AI Hypercomputer setup. Staff interventions to arrange compute resources fell from 20 per week to zero.

SOURCES · cloud.google.com
1 of 3

Upcoming

Announced or reported, but not yet real or available.

Compute & InfraOct 31

Anthropic signs $44.6 billion contract for Nscale's first campus building

Waiting on: Construction - The contracted first building is part of Nscale's planned West Virginia campus, which remains under construction.

techcrunch.com
Compute & InfraOct 29

Fujifilm plans fivefold expansion of magnetic-tape production

Waiting on: Factory expansion - Fujifilm's investment project is scheduled to run from fiscal 2026 through fiscal 2028.

arstechnica.com
TextOct 30

Apple schedules China AI features for iOS 27.3

Waiting on: Software release - Apple has scheduled the China features for an iOS 27.3 beta on November 4, 2026, followed by a full release in December 2026.

theverge.com
↑ Top