From Naptime to Big Sleep: Using Large Language Models To Catch Vulnerabilities In Real-World Code.

From Naptime to Big Sleep: Using Large Language Models To Catch Vulnerabilities In Real-World Code - Project Zero

Posted by the Big Sleep team Introduction In our previous post, Project Naptime: Evaluating Offensive Security Capabilities of Large Language Model...

#ai #cybersecurity #google

Boring language but interesting White House memo, clearly illiustrates the risks involved with AI and that we are not dealing with some kind of hype or bubble only. "AI has emerged as an era-defining technology". https://www.whitehouse.gov/briefing-room/presidential-actions/2024/10/24/memorandum-on-advancing-the-united-states-leadership-in-artificial-intelligence-harnessing-artificial-intelligence-to-fulfill-national-security-objectives-and-fostering-the-safety-security/ #AI #whitehouse #risks

Nice benchmark for AI models, Simple Bench.

SimpleBench

SimpleBench

#AI #benchmarks #simplebench #llm

Let's pray for a Harris win, if only for the people in Ukraine.

As US votes, Ukraine’s future hangs in balance

Ukrainians on the front lines tell the BBC's James Waterhouse that their country’s survival is in the hands of Americans.

#ukraine #harris #trump #elections

Let the AI search war begin , Open AI joins the fray. Although for paying users only. In the US you have also Google AI enabled search and there is also Perplexity (my favorite) everywhere.

The Verge

OpenAI’s search engine is now live in ChatGPT

It’ll roll out to everyone soon.

#ai #search #chatgpt #google #perplexity

Great must-read blog about this: "The emphasis on human oversight as a protective mechanism allows governments and vendors to have it both ways: they can promote an algorithm by proclaiming how its capabilities exceed those of humans, while simultaneously defending the algorithm and those responsible for it from scrutiny by pointing to the security (supposedly) provided by human oversight."

Pluralistic: AI’s “human in the loop” isn’t (30 Oct 2024) – Pluralistic: Daily links from Cory Doctorow

#AI #humanoversight #humanintheloop

Interesting detail in this Google blog: "We're also using AI internally to improve our coding processes, which is boosting productivity and efficiency. Today, more than a quarter of all new code at Google is generated by AI, then reviewed and accepted by engineers. This helps our engineers do more and move faster."

Google

Q3 earnings call: CEO’s remarks

Our Q3 results were led by great performance in Search, Cloud and YouTube.

#AI #google #code #sundarpichai

Good article about NotebookLM, why it's so great and what you can do with it. #AI #Google #NotebookLM https://www.fastcompany.com/91217096/googles-notebooklm-is-one-of-the-most-powerful-use-cases-for-ai

Can AI help humans find common ground in democratic deliberation? Fascinating stuff because it’s not only about technology but also about humans and society, mediation and deliberation. https://www.science.org/doi/epdf/10.1126/science.adq2852 #AI #habermas #deepmind #mediation #deliberation

Still very early days when it comes to "AI agents" but Anthropic shows some first prototypes. It's to extrapolate a few years from now when smart agents will be able to fully control your computer on your behalf and accomplish all kinds of tasks. Also plenty of risks. #AI #Anthropic #agent

NotebookLM: How Small Teams Can Achieve Outsized Results inside Big Tech.

NotebookLM: How Small Teams Can Achieve Outsized Results inside Big Tech

The inside story of NotebookLM's success, 6 lessons for small teams inside big tech, and 7 ways to get the most out of the product

#ai #notebooklm

Google opensources watermarking technology. https://www.nature.com/articles/s41586-024-08025-4 #ai #opensource #watermark

No surprises here https://www.firstpost.com/tech/apple-intelligence-development-at-least-2-years-behind-industry-leaders-claims-report-13827519.html #apple #ai

In the debat about whether AI / LLMs can reason or not it's to good to remember this quote from 1984 from Dijkstra , a dutch computer scientist, ""The question of whether Machines Can Think... is about as relevant as the question of whether Submarines Can Swim." #AI #LLM #reasoning

"What he discovered seems simple on its surface, but the quality of the result has deeper implications for the future of AI assistants, which may soon be able to see and interact with what we're doing on our computer screens."

Ars Technica

Cheap AI “video scraping” can now extract data from any screen recording

Researcher feeds screen recordings into Gemini to extract accurate information with ease.

#AI #video #scraping

Great summary of the debate about LLMs and whether they are reasoning or not and what their quality is in that respect.

The LLM Reasoning Debate Heats Up

Three recent papers examine the robustness of reasoning and problem-solving in large language models

#AI #reasoning #LLM

LinkedIn uses it's data for training AI models, nothing new, but the update of their privacy policy reminded me about that. It must offer amazing and scary possibilities if you think about the fact that the complete social graph of people and businesses is in LinkedIn data. Also we have no way to independently check/audit that data for EU citizens is excluded (or am I missing something). #linkedin #microsoft #ai #privacy #personaldata #eu

Good essay to read, whether you agree with him or not, well argumented and written by Dario Amodei from Anthropic. "Machines of Loving Grace, How AI Could Transform the World for the Better"

Dario Amodei — Machines of Loving Grace

How AI Could Transform the World for the Better

#AI #future #darioamodei #anthropic

Very powerful opensource model released by Nvidia. #AI #Nvidia #opensource #nemotron

nvidia/Llama-3.1-Nemotron-70B-Instruct · Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

NotebookLM experimentje gedaan, toespraak van WIlhelmina voor Radio Oranje ( 28-11-1944) geupload (geluidsfile, 197 MB). Na 2 minuten analyse van de toespraak, beschikbaar in Notebook, met een samenvattig etc, als ik er naar vraag ook een letterlijke transcriptie (!). De potentie van de deze tool voor analyse van content, ook voor audio, is enorm. En dan heb ik het nog niet eens over de veelgenoemde podcast functionaliteit. #AI #NotebookLM #audio