Elimufy Logo Elimufy

10/08/2023 10:47 PM 1054

How Hackers and Ordinary People are Making AI Safer

Artificial intelligence (AI) is advancing at a breathtaking pace. Systems like GPT-4 can generate human-like text, Google's Imagen can create photorealistic images from text prompts, and tools like DALL-E 2 can conjure up fantastical digital art. However, with such rapid progress comes significant risks. Recent examples like bots programmed to imitate real people or AI-generated fake media have raised alarms about potential harms if AI systems are misused or act unpredictably. This has led to the emergence of "red teaming" in AI - an approach where researchers deliberately try to find flaws and vulnerabilities in AI systems before bad actors can exploit them. My blog post will explore the rise of red teaming, how it is making AI safer, and the challenges ahead.

The Roots of Red Teaming

Red teaming has its origins in military strategy, where one group would roleplay as "opposing forces" to test vulnerabilities in operations or technology. The concept has since expanded into the corporate world and now the tech industry. Google, Microsoft, Tesla and other leading companies use red teams to hack their own products and find security holes. The idea is simple - discover problems before hackers in the real world do. Red teaming has mostly been an internal exercise, with employees probing their own systems. But now, tech firms are inviting external hackers and researchers to put AI systems to the test through organized "Generative Red Team Challenges."

Uncovering Flaws Before They Become Real-World Threats

In August 2023, an inaugural generative red team challenge focused specifically on AI language models was held at Howard University. This event, covered by the Washington Post, involved hackers trying to make chatbots malfunction or behave in dangerous ways. For instance, one bot fabricated a completely fictitious story about a celebrity committing murder. While shocking, this demonstrates the need for scrutiny before AI systems interact with real humans. The event was a precursor to a larger public contest at the famous Def Con hacking conference in Las Vegas.

At Def Con's Generative Red Team Challenge, organized by Anthropic with support from the White House, elite hackers went up against the latest natural language AI models from companies like Google, OpenAI, Anthropic and Stability. They were tasked with uncovering flaws and vulnerabilities however possible. Previous internal red teaming by OpenAI revealed risks like GPT-3's potential to help generate phishing emails. The results at Def Con will be kept secret temporarily so the issues can be addressed. But the exercise underscores how seriously developers are taking AI safety amid rising public concerns.



Government bodies like the National Institute of Standards and Technology (NIST) have also conducted controlled testing environments inviting external hackers, researchers and ordinary users to experiment with AI systems. The goal is to discover undesirable behavior or deception before deployment. For instance, in 2020, NIST tested facial recognition algorithms from dozens of companies for accuracy and bias. It found higher error rates for Asian and African faces, demonstrating the need for more diverse training data. Red teaming is increasingly seen as crucial for flagging such problems early when they are easier to fix.

Potential Harms Beyond Just "Hacks"

However, the dangers of AI systems involve more than just direct hacking, security flaws or getting tricked into falsehoods. As pointed out by Rumman Chowdhury of Humane Intelligence, there are also "embedded harms" to look out for. For example, biases and unfair assumptions baked into the AI's training data or the creators' own cognitive biases. Historical data reflects existing discrimination and imbalances of power, which could get perpetuated through AI systems. 

Issues around fairness, accountability and transparency are hard to uncover through technical hacking alone. They require input from diverse communities and viewpoints. Initiatives like Google's Human-AI Community offer platforms for public discussion and feedback around AI development. There are also emerging startups like Scale AI that provide 'bias bounties' - incentivizing ordinary users from different backgrounds to interact with AI and uncover harms. 

Challenges of Scaling and Implementation

Red teaming exercises have shown immense promise in strengthening the safety and reliability of AI before deployment. But there are challenges too. Firstly, there is the issue of scale. Can enough vulnerabilities be identified given the rapid pace of evolution? The parameters and use cases are practically infinite. Tech policy expert Jack Clarke highlights that red teaming needs to occur continuously, not just before product launch. 

Secondly, there is the question of implementation. Identifying flaws is the first step - patching them is equally critical but difficult. Take the recent case where an Anthropic researcher got Claude, the company's AI assistant, to make up scientifically plausible but harmful claims around plastic pollution. While concerning, fixing this requires significant retraining. There is an art to tweaking models without compromising performance.



Lastly, striking a balance between openness and secrecy around red team events is important but tricky. Being transparent about the shortcomings found builds public trust. But excessive openness allows bad actors to weaponize the discoveries before solutions are implemented. The delayed public release of red team results is an attempt to balance these needs.

The Path Ahead

Red teaming provides a proactive way for AI developers to stay ahead of adversaries and mitigate risks preemptively. While not foolproof, it is a powerful paradigm and its popularity will only grow as AI becomes more pervasive. Going forward, the involvement of policymakers and the public along with internal testing will be key to making these exercises more robust and meaningful. Initiatives like the Generative Red Team Challenge, guided by multi-stakeholder participation, point the way towards safer and more beneficial AI for all.

The tech industry still has a lot to prove regarding AI safety. But the commitment shown by leading firms to voluntary red teaming and external scrutiny demonstrates responsible steps in the right direction. AI has immense potential for improving human lives. With care and diligence, we can develop this rapidly evolving technology in sync with shared ethical values. Red teaming powered by diverse viewpoints offers a promising path ahead amid the AI revolution.

You might also interested

30/08/23

Unleash the Power of AI: How to Use ChatGPT on Telegram

The rise of artificial intelligence (AI) is transforming how we access and use information. AI-powered chatbots and virtual assistants like ChatGPT have made obtaining answers and knowledge easier than ever before. In this blog, we will explore how the platform Elimufy allows you to harness the power of advanced AI through seamless integration with the popular messaging app Telegram. You will learn what Elimufy and ChatGPT are, the benefits of accessing Elimufy via Telegram, how to set up and use Elimufy on Telegram, and how AI assistants can enhance learning and creativity. We discuss the future potential of AI in education and how services like Elimufy make AI accessible to all. Whether you want to level up your knowledge or build the next big idea, this blog provides insights on how to unleash AI to get the most out of your pursuits. Read on to discover how you can tap into this revolutionary technology today using Elimufy and Telegram.

Read more

18/10/23

What is AI? Demystifying Artificial Intelligence

Let's take a fascinating journey together, plunging into the world of Artificial Intelligence (AI). You've probably heard about AI changing the world around us, but what is it really? How does it work? From its humble beginnings to the complex technology that it is today, we're going to break it all down for you. We'll explore how different elements like machine learning and big data work together to make AI a reality. And, it doesn't stop there. We'll also examine how AI is shaping various industries and look at what the future holds. However, every coin has two sides, and so does AI – we'll discuss the challenges we need to overcome. So, if you've been curious about AI and looking for a straightforward, jargon-free explanation, you're in the right place!

Read more

27/09/23

The Advantages of Hiring Inexperienced Workers in the AI Era

In a rapidly evolving job market where artificial intelligence and automation are reshaping work as we know it, the premium placed on experience might well be outdated. Soft skills like communication, adaptability, and emotional intelligence are gaining prominence, as these are attributes machines can't replicate. However, companies continue to stress work experience, even for entry-level positions. This blog post questions this prevailing belief by showcasing the advantages and untapped potential of hiring inexperienced workers.

Read more

28/08/23

20 AI Tools to Boost Your Productivity in 2023

AI is automating rote work, generating insights, and allowing knowledge workers to focus on creative, strategic tasks. Tools like 12ft unlock paid content, Photoroom creates ecommerce images, Mayday optimizes calendars, Recall answers questions, and Stylized draws objects. Tugan writes emails, Pico builds web apps, Xembly manages work, and Claid edits backgrounds. Bardeen automates tasks, Onesta answers finance questions, and ChatGPT Writer composes messages. Additional tools build websites, sort photos, assist teachers, generate art, edit video, write emails, translate videos, organize work, and more. AI will be a gamechanger for productivity in 2023.

Read more

15/08/23

The Art of Show vs Tell: Crafting Effective Prompts for Generative AI

ChatGPT and other generative AI systems have captured the public's fascination recently. But behind these tools lies the critical art of crafting effective prompts. A well-designed prompt acts like a genie's lamp, guiding the AI to deliver useful, relevant responses. Poor prompts lead to nonsensical or biased output. In this post, we explore how the timeless writing principle of "show vs tell" can help create better AI prompts. Show-me prompts demonstrate the desired output through examples, while tell-me prompts explain specifications directly. Both approaches have tradeoffs. While generative AI continues advancing rapidly, thoughtful prompt design will remain key to steering these models safely and effectively.

Read more

27/09/23

ChatGPT Advances with Voice and Image Capabilities

In an innovative leap, OpenAI's AI assistant, ChatGPT, has recently incorporated next-level voice and image functionalities. Poised for a roll-out within the next two weeks to Plus and Enterprise users across all platforms, these path-breaking enhancements promise a more engaging and intuitive user interface. The voice capabilities facilitate genuine back-and-forth voice conversations, while the image recognition feature enables the AI to converse about the contents of any given photo. This blog post delves into how these key advancements empower users to seamlessly integrate AI into everyday tasks, the inherent challenges of their implementation, and the steps OpenAI is taking to ensure a safe, effective, and gradual deployment.

Read more