Artificial Intelligence Ecosystems

Even while data professionals ‘seem’ to understand the many challenges of building new ML/ AI tools, they often ignore them while implementing They talk about data quality, business needs, engineering, etc. but then forget about it two weeks into the project On the back of yesterday’s post and my article (link in the comments), here is how you should think about implementing a holistic ecosystem approach for your ML/ AI solutions: 𝟭. 𝗦𝘁𝗮𝗿𝘁 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝘆 🎯 - Define the "Why": Identify specific business problems ML/AI will solve with measurable outcomes - Prioritise Use Cases: Focus on highest business value while considering ecosystem readiness - Secure Executive Commitment: Ensure leadership understands potential AND foundational work - Set Realistic Expectations: Be honest about timelines rather than promising overnight transformation 𝟮. 𝗔𝘀𝘀𝗲𝘀𝘀 𝗬𝗼𝘂𝗿 𝗘𝗰𝗼𝘀𝘆𝘀𝘁𝗲𝗺 𝗥𝗲𝗮𝗱𝗶𝗻𝗲𝘀𝘀 🔍 - Data Foundations & Infrastructure: Evaluate quality/availability of data for priority use cases - Talent and Skills: Map required capabilities against your current team composition - Process Maturity: Can your governance and operational practices support ML/AI deployment? 𝟯. 𝗕𝗮𝗹𝗮𝗻𝗰𝗲 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻 𝗕𝘂𝗶𝗹𝗱𝗶𝗻𝗴 𝘄𝗶𝘁𝗵 𝗣𝗿𝗼𝗴𝗿𝗲𝘀𝘀 🏗️ - Target Foundational Improvements: Strengthen specific components enabling priority use cases - Implement in Phases: Break initiatives into smaller chunks delivering incremental value - Establish Feedback Loops: Regularly evaluate both ML/AI outcomes and ecosystem health 𝟰. 𝗘𝗻𝘀𝘂𝗿𝗲 𝗢𝗿𝗴𝗮𝗻𝗶𝘀𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗘𝗻𝗮𝗯𝗹𝗲𝗺𝗲𝗻𝘁 🤝 - Cross-functional Collaboration: Build frameworks for how teams work together - Continue Investing in Skills: Required capabilities will change across the entire organization - Manage Change: Without stakeholder buy-in, even perfect solutions go unused - Evolve Org Structure & Operating Model: Update how the organization works to reflect AI integration Whenever I hire somebody, I look for their ability to think with a holistic perspective. If you nail this and approach things in this way, you will be much more successful in your data projects and your career! Check out the article (link in the comments) and let me know what you think!

UNESCO for the People – Driving Ethical and Inclusive AI for Humanity Artificial Intelligence is transforming our world. It shapes how we learn, work, and govern – yet billions of people remain excluded from its benefits. At the same time, the risks are mounting: biased systems, opaque algorithms, growing inequalities, and job displacement. This is not only a technological challenge; it is a human rights challenge.   UNESCO has taken the lead by adopting the first global Recommendation on the Ethics of AI – a landmark framework establishing universal principles for fairness, transparency, and accountability. But adoption is only the beginning. The real challenge is inclusive, equitable implementation: turning principles into action so AI serves humanity, not the other way around. At the UNESCO Global Forum on the Ethics of AI in June, scientists, policymakers, and innovators delivered a clear message: ethical AI cannot exist without strong investment in education, infrastructure, and global cooperation.   Throughout my campaign, one lesson stood out: AI must serve people – but first, we must imagine the societies we want, before technology decides for us. “UNESCO for the People” envisions a future where AI promotes peace, equity, and sustainability. Acting with courage, knowledge, and cooperation, we can make AI humanity’s greatest ally by: •Supporting Member States in implementing the 2021 Recommendation on the Ethics of AI, the UNGA resolution adopted in March 2024 on “Seizing the opportunities of safe, secure, and trustworthy AI systems for sustainable development,” and the Pact for the Future. This includes embedding human rights into AI governance so that every system upholds human dignity, freedom of expression, non-discrimination, social justice, international law, and respect for cultural diversity. •Reducing disparities by supporting developing countries through knowledge-sharing, capacity-building programs, innovative financing mechanisms, and the development of infrastructure, multilingual AI systems, and open educational resources – ensuring no community is left behind. • Fostering international solidarity through inclusive dialogue and joint research initiatives that unite governments, academia, industry, and civil society, while promoting human-centered and sustainable AI, rooted in open science. • Making AI a driver of inclusion by leveraging its potential in education, teacher training, youth engagement, local innovation ecosystems, and cultural heritage management. • Anticipating future challenges through a Global Foresight Mechanism to monitor technological trends and prepare societies for their implications, while developing ethical frameworks for frontier technologies such as neurotechnology, quantum sciences, and synthetic biology – ensuring a balance between risks and opportunities before risks outpace regulation.

#AiDays2025 Round Table : #Community Sourcing for low resource languages In an era where AI is fast shaping the contours of our digital future, VISWAM.AI initiative stands as a timely and transformational one. Their mission to build community-sourced Large Language Models (LLMs), grounded in India’s rich linguistic and cultural diversity, is not just pioneering—it’s redefining how inclusive and ethical AI should be built. By anchoring their work in community participation, linguistic preservation, and ethical co-creation, Viswam.ai offers a people-first approach to AI—moving beyond data extraction to cultural stewardship. Their ambition to mobilize 1 lakh community interns to collect data from underrepresented geographies across India is both bold and brilliant. This isn’t just about building better AI—it’s about building equity, agency, and cultural resilience through AI. 1. Linguistic Equity by Design In India, where linguistic hegemony often privileges English and Hindi, AI systems risk reinforcing this imbalance. The solution? Intentional design. Allocate equal engineering and validation efforts to low-resource languages. Ethical AI must be built on informed consent, community ownership, and fair compensation—because data is not just input, it’s identity and heritage. 2. Decentralized Internship Model By decentralizing AI development, we bridge the urban-rural digital divide. This model should focus on: Capacity building through training in ethics and digital literacy Inclusivity by involving women, Dalit and Adivasi youth Localized platforms using mobile-first tools in native languages Partnerships with Swecha, local NGOs, and institutions serve as trust bridges to ensure mentorship and sustainability. 3. Tools for Low-Resource Languages Many Indian languages are oral-first, with complex dialects and sparse corpora. Community-driven solutions—like collecting voice datasets from folklore, and crowdsourcing annotation—are key. Elders, poets, and storytellers become linguistic technologists, preserving not just language but legacy. 4. Trust & Transparency Bias in AI is structural. To mitigate it: Include diverse dialects and accents in training Conduct bias testing and community validation Promote explainable AI with local language dashboards and storytelling What’s Next? A living white paper on ethics, governance, and technical guidelines A roadmap for the internship program, with toolkits and impact metrics Collaboration with literary and linguistic organizations to enrich model depth VISWAM.AI is planting seeds for an AI movement rooted in language justice, data sovereignty, and community wisdom. Let’s co-create systems that don’t just understand our languages—but respect our voices. DC* Chaitanya Chokkareddy Kiran Chandra Ramesh Loganathan Centific

AI in Education is not just plug-and-play 🚸 To ensure a consistent, evidence-based, and inclusive adoption of AI at the systemic level, several key actions are crucial. First, establishing agile and collaborative governance frameworks is essential, which includes defining clear national AI strategies and policies aligned with education goals, while updating existing guidelines to frameworks such as the OECD AI framework. Piloting AI interventions, especially through rigorous evaluation methods like randomized controlled trials, is vital for gathering evidence on effective practices and informing data-driven policymaking, while also researching the causes behind past pilot failures. Teacher training and capacity building are critical for ensuring educators possess the digital skills required to responsibly integrate AI and leverage it as a "learning partner." Promoting inclusivity involves addressing the digital divide by guaranteeing equitable access to technology and quality data, particularly in rural and underserved areas. Additionally, fostering critical thinking and AI literacy among both students and teachers is necessary to navigate potential biases, misinformation, and ethical concerns. Finally, strengthening collaboration among all stakeholders, including governments, the private sector, civil society, and academia, is key to developing relevant, scalable, and inclusive AI solutions.

If AI is learning from biased data, what happens to candidates with disabilities? The rise of automated hiring tools may be locking out millions, and no one is noticing, because it's silent. AI now scans resumes and analyzes video interviews, and companies are adopting it faster than ever. A late-2023 IBM survey of over 8,500 global IT professionals found that 𝟰𝟮% 𝗼𝗳 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀 𝗮𝗹𝗿𝗲𝗮𝗱𝘆 𝘂𝘀𝗲 𝗔𝗜 𝗶𝗻 𝗿𝗲𝗰𝗿𝘂𝗶𝘁𝗶𝗻𝗴, and 𝗮𝗻𝗼𝘁𝗵𝗲𝗿 𝟰𝟬% 𝗮𝗿𝗲 𝗰𝗼𝗻𝘀𝗶𝗱𝗲𝗿𝗶𝗻𝗴 𝗶𝘁. The hope was that AI would reduce hiring bias. But in many cases, the opposite is happening. When trained on data that excludes people with disabilities, it learns to overlook them, too. In June 2025, the New York City Bar Association released a report on The Impact of the Use of AI on People with Disabilities (linked in the comments). Findings show that the statistical nature of AI often leads to discrimination, especially against people with disabilities who fall outside the "average" profiles these systems are built around. The scale of the issue is hard to ignore. Some might argue that one biased hiring manager could affect dozens of candidates in a year. But, as Hilke Schellmann points out, a flawed algorithm deployed across a major employer could impact hundreds of thousands. And because many vendors are rushing underdeveloped tools to market (driven by demand and profit, of course), there's little transparency or accountability. Companies using them often avoid admitting potential harm, fearing legal risk. So what can be done? Making AI inclusive requires a complete shift in how it's developed and implemented, with disability inclusion embedded from the start: ▶️ Use better data. Train AI using datasets that reflect the full range of human experiences, including physical, sensory, cognitive, and mental health disabilities, collected ethically and with consent. ▶️ Design with accessibility in mind. Build tools that work for everyone from the beginning. That includes compatibility with screen readers, voice recognition, and adjustable visual environments and formats. ▶️ Co-create with disabled people. Involve people with disabilities at every stage, from ideation to testing to launch. Feedback should be continuous, not one-off. ▶️ Test for bias. Run regular audits to detect and address bias. Create clear pathways for users to report issues and request improvements. One promising tool is the Conditional Demographic Disparity test, co-developed in 2020 by Sandra Wachter, Professor of Technology and Regulation at the University of Oxford. This public framework helps detect bias in hiring algorithms and pinpoint decision criteria driving inequality - enabling fairer, more accurate systems. Amazon and IBM are already using it. Be honest - how confident are we in the tools we're using to screen talent? #InclusiveHiring #HiringBias #AIRegulation #DisabilityInclusion

Last week, as I was excited to head to #Afrotech, I participated in the viral challenge where people ask #ChatGPT to create a picture of them based on what it knows. The first result? A white woman. As a Black woman, this moment hit hard—it was a clear reminder of just how far AI systems still need to go to truly reflect the diversity of humanity. It took FOUR iterations for the AI to get my picture right. Each incorrect attempt underscored the importance of intentional inclusion and the dangers of relying on systems that don’t account for everyone. I shared this experience with my MBA class on Innovation Through Inclusion this week. Their reaction mirrored mine: shock and concern. It reminded us of other glaring examples of #AIbias— like the soap dispensers that fail to detect darker skin tones, leaving many of us without access to something as basic as hand soap. These aren’t just technical oversights; they reflect who is (and isn’t) at the table when AI is designed. AI has immense power to transform our lives, but if it’s not inclusive, it risks amplifying the very biases we seek to dismantle. 💡 3 Ways You Can Encourage More Responsible AI in Your Industry: 1️⃣ Diverse Teams Matter: Advocate for diversity in the teams designing and testing AI technologies. Representation leads to innovation and reduces blind spots. 2️⃣ Bias Audits: Push for regular AI audits to identify and address inequities. Ask: Who is the AI working for—and who is it failing? 3️⃣ Inclusive Training Data: Insist that the data used to train AI reflects the full spectrum of human diversity, ensuring that systems work equitably for everyone. This isn’t just about fixing mistakes; it’s about building a future where technology serves us all equally. Let’s commit to making responsible AI a priority in our workplaces, industries, and communities. Have you encountered issues like this in your field? Let’s talk about what we can do to push for change. ⬇️ #ResponsibleAI #Inclusion #DiversityInTech #Leadership #InnovationThroughInclusion

This is a vulnerable post, so be kind — but this topic feels too important to ignore. The other day, Jennifer Hutchings, asked me about bias in AI. I shared with her this story. Look, I know bias in AI exists, but I’d never felt it as profoundly as when I decided to create AI-generated headshots of myself. After uploading nearly 50 recent, full body photos, I was stunned by the results. The AI-generated images presented a version of my body that felt unrecognizable. Not just noticeably, but drastically slimmer than the pictures I provided. As a female and as a mother to a daughter, this left me very concerned. I started asking myself: Is this what AI “thinks” I should look like? Is this AI’s “standard” of beauty? Is this what AI “thinks” women should look like? Is this the unachievable “Barbie” of the next generation? Sure, we could blame bad technology but that is just masking the real issue: bias tools can and will lead to negative outcomes – much bigger the impacting a person’s self-esteem. I want to be part of the change. Here are a few practical ideas. I would love to hear your ideas as well. - Select AI Partners That Prioritize Diversity and Inclusivity: When choosing AI tools, look for partners who demonstrate a commitment to diversity and ethical practices. Ask about their approach to building inclusive teams, training data, and bias testing. Working with organizations that value transparency and inclusivity. - Ensure Your AI Council Reflects a Range of Experiences and Perspectives: Build a council that goes beyond gender and race diversity to include a mix of experiences, body backgrounds, and viewpoints. A council with varied perspectives is more likely to identify and address hidden biases. - Engage Actively to "Train" AI Models for Diverse Perspectives: When using AI tools, prompt them to provide multiple perspectives, challenge underlying assumptions, and apply varied cultural, social, or contextual lenses. Encourage your team to ask questions that uncover alternative viewpoints and push for more inclusive responses. Ashley Gross Liza Adams Patty Parobek Cathy McPhillips Claire du Preez #AIInnovation #AIforGood #EthicalAI #InclusiveAI #WomeninTech

I tried using Meta AI to privately process a short Yoruba message before texting my mom. Instead of refining the text, it simply repeated the same sentence three times. This experience revealed something much larger about the current state of artificial intelligence. Despite the breathtaking progress in large-scale models, AI systems still struggle to understand local context, nuance, and culture. A billion-parameter model trained mostly on English text can simulate intelligence, but it often fails to comprehend meaning where local identity, dialect, or social norms matter most. That experience reaffirmed a conviction I’ve held for some time: the next leap in AI will come not from building ever-larger models, but from building localized small language models (SLMs)—AI systems designed to understand the languages, traditions, and lived realities of the communities they serve. If we want AI that truly benefits people, we must first invest in AI infrastructure for local intelligence. Once that foundation is strong, we can build federated AI infrastructure that connects those localized models—sharing insights, not raw data, across borders and industries in compliance with local and international laws. The countries and institutions that get this right will lead the next wave of AI innovation. The future of AI will not just be intelligent—it will be locally fluent, culturally aware, and globally connected. #LocalAIGov #SLMs #FederatedLearning #LanguageTech

AI is only as inclusive as the voices driving its development. The way we build and implement AI today will determine how it serves tomorrow. The choice is ours. It has the potential to reshape industries, but if left unchecked, it risks deepening societal divides and the inclusion gap. While we see progress. Western-centric AI development has perpetuated biases by relying on incomplete data and overlooking underserved regions. To shift this narrative, we need to move beyond the buzzwords and focus on tangible actions. Here’s how: → Diversify the data: We must actively collect and incorporate data from underrepresented regions, ensuring AI systems reflect diverse needs and experiences. → Empower diverse talent: AI development must include voices from all communities. We need initiatives that nurture talent in underserved populations to bring fresh perspectives into tech. → Engage globally: Policymakers, tech companies, and healthcare providers must collaborate, ensuring AI solutions are designed for global accessibility. → Hold ourselves accountable: Regular audits for bias in AI systems should become the norm. → Rethink governance: We need inclusive AI governance that prioritizes representation, particularly when it comes to health and social welfare. → Learn from local experts: Before implementing AI in new regions, tech developers must work alongside local experts to understand cultural nuances and real-world needs. Moreover, applying the 4D Framework: Develop, De-identify, Decipher, De-bias. We can create AI systems that are not just smarter, but also fairer, more inclusive, and global. It’s time to change the conversation. But this isn't just about building better tech. It's about expanding access, education, and funding to communities that have been left behind. It’s about ensuring that every person, no matter where they live, has a seat at the table. AI’s future doesn’t belong to one group. It belongs to all of us. The real question is: Will we design it for everyone?

💡AI isn’t just a tech revolution. It’s a values test. Because if we’re not careful, the AI “advantage” becomes just another privilege some get by birth, proximity, or budget. While we’re hosting keynotes, building agents, and debating prompt frameworks…. millions of brilliant minds still lack access to AI tools, training, infrastructure—or even a seat at the table. That’s not innovation. That’s exclusion with better branding. So here’s what I’m asking companies, leaders, and AI evangelists: • Are you only building for people who already have access? • Is your AI strategy inclusive of rural talent, frontline workers, older employees, or communities historically left out of tech revolutions? • Have you audited your AI training data for bias—and your AI enablement plans for gatekeeping? This movement is really about building a future where only a few get to build at all. So what can we actually do? • Fund access: AI tools, not just licenses. Community hubs. • Build in local context: Train models for language, culture, and use cases outside the tech bubble. • Democratize training: Pay people to learn AI. Certify them. Coach them. • Reward inclusive design: Add equity to your success metrics, not just speed or savings. • Co-create: Bring those impacted by AI into your build process—early and often. If AI is meant to “scale humans,” let’s make sure all humans are invited to scale. Not just the ones with fancy degrees, high-speed Wi-Fi, and an invite to the beta. – LC Just a thought.