My AI Quest’s Post

🚨 AI Training Gone Wrong: TikTok's Parent Company Fires Intern Over $10M Incident Did you hear about the ByteDance intern who allegedly disrupted an AI model's training, causing a stir in the tech world? This incident highlights the critical importance of AI security and the need for stringent protocols in AI development. As AI becomes increasingly central to business operations, companies must prioritize safeguarding their AI models and data. This event serves as a wake-up call for organizations to reassess their AI training processes and access controls. 💡 Pro Tip: Implement a robust permission system and conduct regular security audits for your AI projects. This can help prevent unauthorized access and potential disruptions to your AI development efforts. Despite this setback, ByteDance remains a leader in AI innovation, particularly with its popular chatbot Doubao. It's a reminder that even tech giants face challenges in the rapidly evolving AI landscape. What measures does your company take to protect its AI assets? Share your thoughts or experiences below! #AIInnovation #TechSecurity #FutureOfWork https://lnkd.in/d9ANYiTB

Ethics in AI: Addressing Challenges and Ensuring Responsible Technology Development Responsible AI practices prioritize ethical considerations throughout the development lifecycle to ensure that AI technologies are used for the greater good and do not perpetuate harm or discrimination. AI algorithms may exhibit bias, leading to unfair outcomes and perpetuating existing societal inequalities. Responsible AI enables the design, development and deployment of ethical AI systems and solutions. Ethical AI acts as intended, fosters moral values and enables human accountability and understanding. # TalentServe Internship

Your LinkedIn activity is shaping AI, and you might not even know it. Here's what you need to know: LinkedIn is using our data for various AI purposes, including writing assistant features. Thankfully, they've introduced a new privacy setting that allows users to opt out of AI model training. Verge says opting out doesn't affect training that has already taken place. To protect your data, you'll need to take 2 steps: i. Go to your account settings, find "Data privacy" tab, and turn off "Data for Generative AI Improvement." ii. Fill out the LinkedIn Data Processing Objection Form to opt out of other machine learning tools. LinkedIn is using privacy-enhancing technologies to protect personal data in their training sets. This move follows a similar admission by Meta about scraping user data for AI training. It's a stark reminder that our online presence is increasingly becoming fodder for AI development. What are your thoughts on this? #linkedin #privacy #AI

Do you use AI tools to increase your productivity for at least 10% of your work activities? If no. Then YES! Take the course. [And, maybe take it even if you do.] MY CONTEXT: I’m looking for an internship. Yes. A marketing internship [click through the link in my profile to see my pitch]. And, I want to skill-up to land the best experience out there. One of the action items from my first [awesome] interview was to take an AI course. Google AI Essentials seemed like the best starting point. COURSE SPECIFICS: 10 hours. 5 modules. Available through Coursera. COURSE CONTENT: Videos from Google AI Execs who are on the leading edge of AI development. Practical guided activities in the Google Workspace. And, 5-question Challenges at the end of each module. MY OPINION: Take the course. I have been a heavy consumer of AI media and lore for a couple of years. And, I am a regular user of ChatGPT. But, I don’t have any subscription services. This course really built a bridge between my foundational understanding of AI and the opportunities for practical application. A bonus of the course is that it unlocks Gemini in your Google Workspace. So, now my work activities and AI assistance will be side-by-side in the same platform. The course also covers AI bias and our responsibility to be dedicated critical thinkers as we use the tool to create content and outcomes that are equitable to all. And, it touches on future learning. You pretty much have to make a written commitment to future learning within and outside of the Google context as you exit the course. Next, I am onto a 50-hour marketing-specific AI course. I will let you know how that goes. Please DM me if you have any questions!

I want your opinion - is this a bad thing? LinkedIn has recently introduced a new setting that allows users to opt-in or out of their data being used to train generative AI models. While this is a significant step towards transparency and user control, it raises important questions about the balance between innovation and privacy. Pros of Sharing Data: - Enhanced AI Capabilities: By leveraging a vast dataset of user-generated content, LinkedIn can train more sophisticated AI models that produce higher-quality, relevant content. - Personalized Experiences: This could lead to more tailored recommendations, job suggestions, and networking opportunities. - Advancement of Technology: Contributing data to AI research can help accelerate the development of beneficial AI applications. Cons of Sharing Data: - Privacy Concerns: There is a risk that personal information could be misused or exposed, potentially leading to identity theft or other harmful consequences. - Algorithmic Bias: If the training data is biased, the AI models may perpetuate existing inequalities or stereotypes. - Lack of Control: Once data is shared, it becomes difficult to control its usage and potential outcomes. What's Your Take? As users, we have a crucial role to play in shaping the future of AI. Do you believe the benefits of sharing data for generative AI outweigh the risks? Or do you prioritize privacy and opt-out of such training? Let's discuss the implications of this decision and explore ways to ensure that AI development is both innovative and ethical. #AI #Privacy #Data #LinkedIn #Technology #Innovation #Ethics

Culture of AI We had a company meeting last week and I challenged all Breach Secure Now employees to adopt AI. Here are some of the messages that I shared with the team: Don’t be afraid to adopt AI We have been talking about AI and many employees have been using AI. Don’t be afraid to say “ChatGPT helped me with this article or with this report or with this analysis”. Share your success stories If you find a good use case for AI, share it. Share your wins in staff meetings. Share new techniques with colleagues. Show them how you leverage this new technology to make your job better. Use it every day Set a goal to use AI every day. The more you use it, the better you get at it. And the better the results will be. It takes hours of practice to form a habit, start today. Think of AI as an intern I use the analogy that AI is like an intern. Interns don’t take your job. Interns make your job easier. But interns need to be trained. You need to invest time to make sure the intern knows how they can help you. This is true for your new AI intern. Take the time to learn how to work with your new intern, and the results could be amazing. The conversation that followed was lively and many employees started to share how they are already using AI. They felt like they were given the freedom now to share their success. I suspect many of your employees or your clients’ employees are using AI but afraid to tell anyone. Developing a culture of AI adoption will be critical to success. Just like a culture of cybersecurity is critical to protecting a company. Here are steps that I believe are required in adopting a culture of AI 1. Awareness - knowing the risks and benefits of this new technology. Raising awareness of how AI can benefit employees and a company. Tip: start a grassroots approach and get employees onboard first. 2. Training - teaching employees how to use and leverage these new AI tools in a safe and effective manner. 3. Openness - encouraging a mindset that is open to experimenting with and adopting new AI technologies. 4. Practices - implementing protocols and procedures on how and when AI should be used. This is where policies, procedures and acceptable use guidelines are needed. 5. Support - companies provide resources, such as AI tools and expert guidance, to help employees implement AI solutions. I believe AI technologies will change the way companies operate. But in order to be successful, companies will need a culture of AI adoption. What are your thoughts? 👇👇

LinkedIn is Training AI on Your Data—Here’s How You Can Opt Out LinkedIn recently introduced a privacy setting that automatically opts users into using their data for #GenerativeAI model training without explicit consent. If you’re concerned about privacy, you can easily turn off this setting by navigating to your Data privacy settings and switching off the Data for Generative AI Improvement option. Keep in mind that opting out only affects future data usage, not the data that’s already been collected. As companies like LinkedIn and Meta quietly collect more data for AI development, it’s important to stay informed and protect your personal information. Read more about it - Link in comments. #DataPrivacy #AI #GenerativeAI #LinkedIn #Meta #MachineLearning #PrivacySettings

The process of training an AI model is no different than training an intern to perform certain business tasks : Can the intern learn quickly if shown the ropes? Yes! Can the intern deliver value by end of week 2? Yes! Can the intern make mistakes and have to be occasionally guided? Yes! Can the intern course correct and pivot? Yes Will the intern replace your job? No Is the intern a failure on not delivering the perfect result? No Can the intern do every business task? No Then why treat your AI platform any differently? If you manage your expectations a bit, then maybe AI can truly be a #gamechanger To do AI right, contact #ibm #watsonx #aigovernance Armand Ruiz IBM IBM watsonx.ai Shuang (Sherry) Yu Lana Strazhkova Ayal Steinberg Tanya Dua Brianne Zavala Alan McAdams Erica Yin

“Ethical concerns about AI in education” went viral yesterday. The article that sparked it was lazy clickbait. The article was called "Teachers are using AI to grade essays. But some experts are raising ethical concerns" (article linked in comments) Here are my top 3 issues with the reporting. 🤔 1. The term “Using AI” is great for drumming up controversial news, it is terrible for having productive discussion. 2. Many of the “ethical concerns” can’t be discussed without specifying what “using AI” actually means. 3. Our conversations about AI need to be more precise. The broader the conversation is, the faster it will become polarized. Details 👇 1. “Using AI” ”Using” is a broad term. “AI” is an imprecise term. “Using AI to grade papers” could mean anything from spellcheckers to running the whole thing through ChatGPT, having it write feedback, and never reading the paper. It could also mean a million things in between. Referring vaguely to “using AI” is a tactic that allows everyone to make assumptions about what is happening, and then they click the article for confirmation. It was effectively written for news and social media. But terribly written for productive discussion. Real discussion about AI require details about WHAT “AI” is being “used” and exactly HOW it is being used. ✅ 2. “Ethical Questions” The article mentions ethical questions that one could have about “using AI”. But it doesn't tie them back to actual tasks, tools, or outcomes. The one example of a potential ethical violation is using student writing to train LLMs without the student’s consent. The article mentions that ChatGPT does this. But it doesn’t mention that this function can be turned off by clicking one button. 🤔 The rest of the “ethical considerations about AI” are vague in both the ethical violation they suggest and the “AI” they claim is responsible for it. I am not saying that there are 0 ethical considerations with “using AI”. Just that we need to discuss them in concrete terms. Not just hint at them in way that gets people worked up. 3. Precision Conversations about AI need to be precise. We can't afford to generalize anymore. As it stands, you can say “AI” in a crowded room and watch as the room divides into two opposing armies. “Pro AI” and “Anti AI”. ⚔️ These polarized views are created by the lack of precision we have in our discussions. Because vague terms allow everyone to create definitions that confirm their biases. If we actually want to figure out this “AI thing” we need to talk about: 🔨Specific tools 🔨Specific tasks 🔨Specific concerns Otherwise we will just be living in our own stories about AI. While others live in theirs. This goes for education, translation, and everything else. Now we need details over dogma.