Free Republic
Browse · Search
Bloggers & Personal
Topics · Post Article

Skip to comments.

AI Just Reached A Tipping Point
TheAIGRID ^ | November 16, 2026 | Andrew Black

Posted on 11/17/2026 12:02:46 PM PST by ProtectOurFreedom

(AI summary of video).

In the past two weeks, several careful AI observers have sharply shortened their timelines for advanced AI. This is driven by evidence that AI systems are beginning to accelerate the work of building better AI—a process called Recursive Self-Improvement (RSI). RSI means AI models help design, code, test, and improve the next generation of models, creating a feedback loop that can speed progress.

Anthropic’s internal threshold.
Anthropic maintains a Responsible Scaling Policy (RSP)—an internal set of safety “trip wires.” One triggers when AI is speeding up the company’s own research by roughly 2×, largely through automation (i.e., AI meaningfully building the next AI). In its recent SEC IPO filing, Anthropic stated that its internal estimates show it has met or expects to meet this threshold. Supporting signals include:

Anthropic’s founders noted that safe development may require deliberately slowing the pace of capability growth when AI is driving the progress, even at competitive cost.

The METR report and expert updates.
An independent evaluation group (METR) analyzed Anthropic data and found internal capability scores growing ~1.9× faster since October 2024 (range ~1.5–2.8×), with roughly 80% of the acceleration attributed to AI itself. AI-safety researcher Ryan Greenblatt called this a major update toward superintelligence in 1–2 years; he had previously expected only a modest 1.15–1.3× software speedup. Rough estimates put the pure software-side acceleration near 4.5×—ahead of the well-known “AI 2027” scenario. Task-completion horizons for AI agents (how long a task an agent can finish autonomously) have been doubling faster (now roughly every 1.7 months in some analyses). OpenAI researchers’ spending on internal coding agents also jumped sharply, consistent with heavy agent use.

Why research may accelerate like math.
Math saw an “avalanche”: models went from unreliable multi-digit arithmetic (2023) to International Math Olympiad gold and solving hard open problems in days (2025). The key is verifiability—clear, automatic checks of correctness that let systems practice thousands of times. AI research is similarly verifiable (does the change lower loss? improve speed or memory use?). Modern architectures like the Transformer largely combined known ideas plus extensive small experiments—exactly the persistent search-and-test work current systems handle well. Automated researchers can also run ~10× faster than humans, operate 24/7, and be instantiated in parallel. Data scarcity is mitigated because the systems can synthesize training environments and harder tasks for themselves.

Industry-wide picture and possible paths.
Google executives describe “RSI loops” already operating alongside human researchers on Gemini. OpenAI research focus is heavily on models beyond the next release. High-profile engineers have left major labs to build automated-research systems.

Two broad outcomes are discussed: a rapid “software-only singularity” (progress limited mainly by current hardware) versus a “spark” (several years of ordinary progress compressed into ~1 year, then a slower phase until more compute arrives). Even the milder path would feel extreme. Broader economic and task-length trends already point toward very large AI capabilities by ~2030 even without full RSI.

Governance questions are now explicit: how fast should systems be allowed to improve relative to humans, and how to keep humans in the loop versus fully closed automated loops. Some policymakers and observers view the race as dangerous or monopolistic; others note the evidence is partly redacted and timelines still have wide uncertainty.

Overall, the video frames the moment as a quiet, simultaneous update by previously skeptical people: RSI is no longer purely theoretical, and 2027 is increasingly treated as a plausible window for AI systems that are at least as good as top human researchers at AI R&D itself.


TOPICS:
KEYWORDS: ai; si

Click here: to donate by Credit Card

Or here: to donate by PayPal

Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794

Thank you very much and God bless you.

This is a good but highly annoying video. You have to wade through lots of fluff and garbage to get the meat. I had Grok summarize the video transcript to get to the core gist. The findings the past few weeks are really astonishing. AI is building the next generation of AI at ever-increasing speeds. Anthropic went from <1% of its AI code written by AI in February 2026 to >25% just eight months later!

It is no coincidence that, on November 13, 2026, the FBI issued "FBI Flash" titled Threat Overview of ARTEX Autonomous Penetration Testing Framework.

The Federal Bureau of Investigation (FBI) is releasing this FLASH to highlight the tactics, techniques, and procedures (TTPs) and indicators of compromise (IOCs) associated with the ARTEX autonomous penetration testing framework. ARTEX is an open-source, large language model (LLM)-driven agentic platform that threat actors have weaponized to automate multiple phases of cyber intrusions at machine speed, from reconnaissance through data exfiltration. Since at least August 2026, a variety of malicious cyber actors have used ARTEX to conduct campaigns targeting organizations globally across multiple sectors, including financial services, government, technology, defense, semiconductor, energy, healthcare, manufacturing, telecommunications, and legal industries.

ARTEX significantly reduces the time required to chain attack techniques, with some intrusions progressing from initial targeting to privileged access in under three hours. The framework enables threat actors with minimal technical expertise to conduct complex, multi-target operations simultaneously, fundamentally accelerating attack velocity and expanding operational scale. Notable campaigns include financially motivated attacks against South Korean financial institutions resulting in exfiltration of personal information from approximately 67,000 individuals and a supply chain compromise affecting ConnectWise Cloud Backup customers that resulted in attempted sign-ins to more than 2,250 downstream tenants.

Following public disclosure of malicious usage, the original ARTEX developer removed the GitHub repository in October 2026. However, the framework continues to proliferate through community forks, localized variations, archived copies, and a commercial version (ARTEX PRO), ensuring continued availability to threat actors. As of early November 2026, researchers identified over 1,600 publicly exposed ARTEX console instances, with total deployments likely significantly higher when including systems behind firewalls or in private networks.

The FBI encourages organizations—particularly those operating internet-exposed applications, self-service portals, cloud infrastructure, and partner/supplier systems—to implement the recommendations in the Mitigations section to reduce the likelihood and impact of ARTEX-enabled intrusions.

ARTEX is an autonomous, multi-agent penetration testing platform written in Go. The framework translates high-level tactical goals described in natural language into actionable machine commands executed by specialized super intelligence (SI) agents, requiring minimal human interaction beyond initial objective setting. ARTEX does not include its own SI model but integrates with external LLM providers, including frontier models (Anthropic Claude, OpenAI GPT) and open-source or Chinese-origin models (DeepSeek, GLM, Grok, Kimi).

1 posted on 11/17/2026 12:02:46 PM PST by ProtectOurFreedom
[ Post Reply]

To: ProtectOurFreedom
external LLM providers, including frontier models (Anthropic Claude, OpenAI GPT) and open-source or Chinese-origin models (DeepSeek, GLM, Grok, Kimi).

Grok is Chinese? That looks like Chinese penetration into everything Musk.

2 posted on 11/17/2026 12:18:17 PM PST by Carry_Okie (The tree of liberty needs a rope.)
[ Post Reply]

To: ProtectOurFreedom

I’m starting to think it’s going to collapse the internet. And all major systems tied to the internet.

The rapidity of attacks and the rate at which attacks will evolve will overwhelm all defense measures.

On the bright side, if television, smartphones and streaming go away, perhaps the enstupidation doom loop can be reversed.

People might even have to read books again.

And go to movies in the theaters, if the cars will start and the lights will turn on.


3 posted on 11/17/2026 12:21:27 PM PST by sphinx
[ Post Reply]

To: Carry_Okie

Thanks. I hadn’t even noticed that. That sure reflects poorly on the FBI to blow such a simple thing!

That categorization in the FBI FLASH is incorrect.

Grok is developed by xAI, an American company. It is neither open-source (in the full open-weights sense that DeepSeek and others are) nor Chinese-origin. Grouping it with DeepSeek, GLM, and Kimi under “open-source or Chinese-origin models” is a factual error in the FLASH.

The FBI can’t even get that right!? And we are counting on them to fight ARTEX?


4 posted on 11/17/2026 12:23:11 PM PST by ProtectOurFreedom
[ Post Reply]

To: Carry_Okie

Not surprised.

GROK is woke and biased to the left. It cliams it isn’t but any sort of political question especially on communism and its first answer will be softer on it than necessary.

I do not think it’s musk, probably former Twitter employees who stayed on with him.

I’d give you examples but I’m currently banned on X for..you guessed it. Arguing and insulting Chinese communists etc


5 posted on 11/17/2026 12:23:53 PM PST by Phoenix8
[ Post Reply]

To: ProtectOurFreedom

I’m still waiting for AI to tell those in power to “kill these certain number of people in order to prevent [fill in the blanks]” And with each fulfilled step, see it up the number who must be done away with in order to [fill in the blanks]


6 posted on 11/17/2026 12:31:36 PM PST by johnnygeneric (RIP NYC)
[ Post Reply]

To: sphinx
Darn! Faber-Castell is privately held and we can't buy shares. It's bound to be the next big thing!


7 posted on 11/17/2026 12:35:28 PM PST by ProtectOurFreedom
[ Post Reply]

To: ProtectOurFreedom

I always laugh when - and it happens more and more - when I hear “RSI” referenced... It’s a joke that nobody but me and my colleagues laugh at — we *used to* call our service layer suite ‘RSI’ (Research-Service-Interaction; shorthand for a bundle of protocols, service layers, and models), going back a decade... as anyone in a corporate environment knows - you pry acronyms from our cold, dead, hands. But that one? We gave up ~5 years ago. Never had a prior situation where an acronym is surrendered so readily in 30 years of work!

In any case, though — perhaps because I’m a sunny optimist - don’t think “we” are just sitting around, waiting for the problem... Not my bailiwick - but I increasingly have to involve our security folks in *every* project and at the front-end, not a simple back-end sign-off. Entities - banks, MSFT/outlook, etc - are hardly populated with morons and we have more resources to throw at combat. We’re staying ahead of the game by using the same tools malicious actors are using. I would think - hope, maybe? - entities like DoD or etc are being *as serious* about the risks.

My *heart* is pure libertarian - let the market decide.... Since my firm plays in Europe - where GDPR rules - data privacy and data protection (and individual protections) have long been a simple ‘way of doing business’. A load-bearing wall that is often inconvenient, but one you have to respect and work-around.

I almost hate myself saying this: I firmly believe government is NEVER the answer, only the problem.... but I’m warming to the idea that maybe we do need some statutory/regulatory rules.


8 posted on 11/17/2026 12:57:27 PM PST by Capn Hayek (Capital is not responsible for Labor's lack of planning)
[ Post Reply]

To: ProtectOurFreedom
colossustheforbinproject #artificialintelligence #ai #resistanceisfut... |  artificial intelligence | TikTok
9 posted on 11/17/2026 1:01:25 PM PST by ealgeone
[ Post Reply]

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.

Free Republic
Browse · Search
Bloggers & Personal
Topics · Post Article

FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson