Impact of Generative AI

Generative AI continues to generate excitement, but significant challenges are often overlooked. Reports from respected sources such as Harvard Business Review and Goldman Sachs highlight that current expectations may not align with reality. The technology, while promising, has limitations that need to be acknowledged and addressed. In May, Harvard Business Review discussed "AI's Trust Problem," in June, Goldman Sachs raised doubts about whether the expected $1 trillion in AI investment will deliver substantial returns. Their concern: aside from developer efficiency, there may not be enough value to justify such massive spending, especially in the near term. Jim Covello, Goldman Sachs' head of global equity research, pointed out that replacing low-wage jobs with costly technology contradicts earlier tech transitions, which focused on improving efficiency and affordability. A recent analysis from Planet Money echoes this skepticism, listing “10 reasons why AI may be overrated.” Issues like hallucinations (when AI generates false or misleading information) and declining quality in AI-generated outputs raise concerns about its readiness for widespread use. A study by The Washington Post also examined what people ask AI chatbots about, revealing unexpected trends. Along with common academic assistance, some topics raised ethical and personal concerns. 🔍 Reality check: Generative AI can be impressive but often struggles with accuracy, leading to errors or hallucinations. 💸 Investment risks: Financial experts question the value of massive investments in AI and wonder if the technology will offer enough returns in the short term. 📉 Productivity vs. quality: While AI can increase productivity, particularly in coding, research shows that the quality of AI-generated code is often subpar. 📚 Help with homework: Students turn to AI chatbots for homework help, but concerns arise when AI provides direct answers rather than guidance or learning support. ❓ Personal and sensitive queries: Many chatbot users ask about personal topics, including sex and relationships, which raises ethical questions about privacy and appropriate use. These points serve as a reminder that while generative AI is a powerful tool, it’s important to approach it with realistic expectations and a clear understanding of its current limitations. #GenerativeAI #AIEthics #AIRealityCheck #AIinEducation #TechInvestments #AIProductivity #AIChallenges #AIHomework #AIandSex #AIinConservation #AIFuture #AIHype 

Researchers from Google's DeepMind, Jigsaw, and Google.org units are warning us in a paper that Generative AI is now a significant danger to the trust, safety, and reliability of information ecosystems. From their recent paper, "Generative AI Misuse: A Taxonomy of Tactics and Insights from Real-World Data": "Our findings reveal a prevalence of low-tech, easily accessible misuses by a broad range of actors, often driven by financial or reputational gain. These misuses, while not always overtly malicious, have far-reaching consequences for trust, authenticity, and the integrity of information ecosystems. We have also seen how GenAI amplifies existing threats by lowering barriers to entry and increasing the potency and accessibility of previously costly tactics." And they admit they're likely *undercounting* the problem. We're not talking dangers from some fictional near-to-medium-term AGI. We're talking dangers that the technology *as it exists right now* is creating, and the problem is growing. What are the dangers Generative AI currently poses? 1️⃣ Opinion Manipulation through disinformation, defamation and image cultivation. 2️⃣ Monetization through deepfake commodification, "undressing services," and content farming. 3️⃣ Phishing and Forgery through celebrity ad scams, phishing scams and outright forgery. 4️⃣ Additional techniques involving CSAM, direct cybersecurity attacks, and terrorism/extremism. Generative AI is not only an *environmental* disaster due to its energy and water usage, and not only a cultural disaster because of its theft of copyrighted materials, but also a direct threat to our ability to use the Internet to facilitate exchange of information and facilitate commerce. I highly recommend giving this report a careful read for yourself. #GenerativeAI #Research #Google #Cybersecurity #Deepfakes https://lnkd.in/gR99hZhe

Generative AI (GenAI) has ushered in a renaissance age for the generalist. For years, organizations have spent a disproportionate amount of capital hiring hyper specialized talent with deep technical knowledge. Now, with the democratization of #GenAI, the value offered by hiring ‘capable generalists’ is on the rise. People who articulately frame their thoughts, pose well-formed questions (prompts), and exercise #AI tools to their advantage, stand to benefit greatly. The demand for specialized AI talent - model developers, AI ops talent, and engineers to build and maintain infrastructure - will persist. But demand for non-technical talent is shifting to a more balanced state. Those who have the skills to extract value from platforms are becoming as valuable to organizations as those who build them.  I strongly encourage business leaders to incorporate skills like curiosity, critical thinking, and effective writing into their hiring profiles. These skills are becoming increasingly important - and valuable - in this next phase of technology and operations.

Check out this massive global research study into the use of generative AI involving over 48,000 people in 47 countries - excellent work by KPMG and the University of Melbourne! Key findings: 𝗖𝘂𝗿𝗿𝗲𝗻𝘁 𝗚𝗲𝗻 𝗔𝗜 𝗔𝗱𝗼𝗽𝘁𝗶𝗼𝗻 - 58% of employees intentionally use AI regularly at work (31% weekly/daily) - General-purpose generative AI tools are most common (73% of AI users) - 70% use free public AI tools vs. 42% using employer-provided options - Only 41% of organizations have any policy on generative AI use 𝗧𝗵𝗲 𝗛𝗶𝗱𝗱𝗲𝗻 𝗥𝗶𝘀𝗸 𝗟𝗮𝗻𝗱𝘀𝗰𝗮𝗽𝗲 - 50% of employees admit uploading sensitive company data to public AI - 57% avoid revealing when they use AI or present AI content as their own - 66% rely on AI outputs without critical evaluation - 56% report making mistakes due to AI use 𝗕𝗲𝗻𝗲𝗳𝗶𝘁𝘀 𝘃𝘀. 𝗖𝗼𝗻𝗰𝗲𝗿𝗻𝘀 - Most report performance benefits: efficiency, quality, innovation - But AI creates mixed impacts on workload, stress, and human collaboration - Half use AI instead of collaborating with colleagues - 40% sometimes feel they cannot complete work without AI help 𝗧𝗵𝗲 𝗚𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 𝗚𝗮𝗽 - Only half of organizations offer AI training or responsible use policies - 55% feel adequate safeguards exist for responsible AI use - AI literacy is the strongest predictor of both use and critical engagement 𝗚𝗹𝗼𝗯𝗮𝗹 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 - Countries like India, China, and Nigeria lead global AI adoption - Emerging economies report higher rates of AI literacy (64% vs. 46%) 𝗖𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝗤𝘂𝗲𝘀𝘁𝗶𝗼𝗻𝘀 𝗳𝗼𝗿 𝗟𝗲𝗮𝗱𝗲𝗿𝘀 - Do you have clear policies on appropriate generative AI use? - How are you supporting transparent disclosure of AI use? - What safeguards exist to prevent sensitive data leakage to public AI tools? - Are you providing adequate training on responsible AI use? - How do you balance AI efficiency with maintaining human collaboration? 𝗔𝗰𝘁𝗶𝗼𝗻 𝗜𝘁𝗲𝗺𝘀 𝗳𝗼𝗿 𝗢𝗿𝗴𝗮𝗻𝗶𝘇𝗮𝘁𝗶𝗼𝗻𝘀 - Develop clear generative AI policies and governance frameworks - Invest in AI literacy training focusing on responsible use - Create psychological safety for transparent AI use disclosure - Implement monitoring systems for sensitive data protection - Proactively design workflows that preserve human connection and collaboration 𝗔𝗰𝘁𝗶𝗼𝗻 𝗜𝘁𝗲𝗺𝘀 𝗳𝗼𝗿 𝗜𝗻𝗱𝗶𝘃𝗶𝗱𝘂𝗮𝗹𝘀 - Critically evaluate all AI outputs before using them - Be transparent about your AI tool usage - Learn your organization's AI policies and follow them (if they exist!) - Balance AI efficiency with maintaining your unique human skills You can find the full report here: https://lnkd.in/emvjQnxa All of this is a heavy focus for me within Advisory (AI literacy/fluency, AI policies, responsible & effective use, etc.). Let me know if you'd like to connect and discuss. 🙏 #GenerativeAI #WorkplaceTrends #AIGovernance #DigitalTransformation

It’s easy to think of AI as a time-saver that streamlines workflows and accelerates output. But the deeper opportunity lies in how it’s reshaping the nature of work itself. A new study from Harvard Business School’s Manuel Hoffmann followed more than 50,000 developers over two years, with half using GitHub Copilot. The results were striking: developers shifted away from project management and toward the core work of coding. Not because someone told them to, but because AI made it possible. With less need for coordination, people worked more autonomously. And with time saved, they reinvested in exploration—learning, experimenting, trying new things. What we’re seeing here isn’t just productivity. It’s a shift in how work gets done and who does what. Managers may spend less time supervising and more time contributing directly. Teams become flatter. Hierarchies adapt. This is just one signal of how generative AI is changing our org charts and challenging us to rethink how we structure, support, and lead our teams. The future of work isn’t just faster. It’s more fluid. And if we get this right, it’s a whole lot more human. https://lnkd.in/gaUgXnRY

GenAI Beyond Art, Video, Text: Addressing the World's Challenges   McKinsey & Company Global Institute modelling of trends in AI adoption revealed that AI has the potential to deliver additional global economic output of about $13 trillion by 2030, which would increase GDP by approximately 1.2 percent per year. There are many examples of global AI applications and use by governments to improve social welfare, national health care systems, domestic security and surveillance, and transportation. We have seen during Covid that global interconnectivity for data and treatments is lacking. +Disaster Relief and Infrastructure Development: Generative AI can model natural disasters, generating new patterns to help governments prepare and respond effectively. Real-time text and voice generation ensure efficient communication with affected populations, aiding in disaster relief efforts. +Healthcare: AI applications revolutionize healthcare by diagnosing diseases, recommending treatments, and enhancing patient engagement. Synthetic medical image generation augments datasets, improving diagnostic accuracy. Accelerates drug discovery and molecular design, enabling the development of life-saving medications. Text-generation educates patients effectively. +Education: Generative AI enriches learning experiences by generating quizzes, exercises, and interactive simulations. Personalized learning plans and textbook recommendations. Virtual tutors and language learning companions, powered by image and voice generation, provide adaptive learning experiences. +Wildlife Conservation: With a 69% average reduction in species populations since 1970, generative AI becomes vital. Predicts ecological changes and population dynamics, aiding researchers in creating proactive strategies to protect endangered species. +Financial Inclusion and Human Rights: Generative AI can also contribute to solving challenges in these areas. It helps promote financial inclusion through personalized financial planning and innovative credit scoring models. In human rights, it aids in automated translation, document analysis, and combating online harassment. KOREAN APPROACH TO AI The Korean government released its national strategy for AI on December 17, 2019. The strategy was formed based on the AI ecosystem, AI use, and people-centered AI and consists of 100 government-wide action tasks under nine strategies (Figure 12 see report). With its New Deal strategy, Korea is expected to transform into the smarter country to use data and digital technologies, including AI, and leads the innovative public services. 💜Generative AI's evolving nature and increasing capacity to contribute to global society make it an exciting field. By harnessing its innovative potential, we can address complex challenges, climate, inclusion and create a better future for all of us. What are the use cases you a excited about making a change in your life? #AI #generativeAI #innovation #smartcities #marthaverse

New survey: Internal GenAI tools are booming; client-facing use cases are lagging. Here’s why in 2 words: Hallucinations and PR. Leaders are (rightly) spooked by: • Safety issues with GenAI going off the rails • PR disasters from mishaps (Air Canada 👀) But there's actually a deeper problem here (in my opinion): Risk Management and Governance. Without the right structures, you can't afford to launch client-facing GenAI tools. The risk is too high. You need to be investing in things like: • Dedicated AI governance committee • Company-wide AI ethics & code of conduct • Complete, operationalised AI risk management frameworks • Robust data governance policies for quality, provenance, privacy, and security • Controls, audits and risk assessments of AI systems, including third-party tools Speaking with leaders over the last 12 months, most organisations are far behind here. We need to get moving - fast. Because right now, AI innovation isn't limited by capability or compute power. It's limited by poor risk management and governance. Until we bed that down, GenAI will just be a shiny tool for cost-cutting and efficiency - not a tool for transforming how products and services are delivered. -- PS. What do you think? Do you agree that risk management and governacne are issues here? Or is something else going on? Would love to hear your thoughts below.

AI Economics: Bold Predictions, Harsh Realities Remember PwC's bold prediction that AI would contribute $15.7 trillion to the global economy by 2030? Meanwhile, the current reality portrays a different picture. An MIT study reveals that 95% of corporate GenAI pilots are failing. Also, the FT’s three-part AI series highlights ballooning capex, power constraints, and shaky unit economics around data centers. And the AI Index 2025 notes uneven productivity gains. Let's pause and examine what's actually happening on the ground. The MIT study reveals that 95% of corporate generative AI pilots are failing to deliver measurable business impact. Despite $30-40 billion in enterprise AI investment, only 5% of initiatives achieve rapid revenue acceleration. The culprit isn't the technology; it's flawed integration and misalignment with existing workflows. This reality check comes as various reports, including analyses from major financial publications, highlight the growing disconnect between AI promises and practical outcomes. We're witnessing "GenAI Divide", a stark gap between expectations and execution. The path forward, in my opinion, requires honest recalibration: ✔️ Start small, think workflow-first: Integrate AI into existing processes rather than forcing wholesale changes ✔️ Measure what matters: Define clear success metrics beyond tech demos; focus on P&L impact ✔️ Invest in change management: 95% failure rate suggests this is more about people and processes than algorithms  ✔️ Build gradually: Successful companies are treating AI as a marathon, not a sprint ✔️ Ship safely: policy, auditability, and human-in-the-loop by default. The trillion-dollar AI revolution might still happen, but it won't be through blind faith in shiny pilots. It'll come from organizations that approach AI with strategic patience, clear objectives, and ruthless focus on real-world value creation. Ambition is good. But disciplined execution, not hype, will determine who captures real AI value. #AI #TechReality #InflatedExpectations

Generative AI is showing incredible promise in proof-of-concept (PoC) environments. But as we move toward production, the real complexities are coming up - and one of the biggest is user consent. In the PoC stage, we often work with sanitized or synthetic data. In production, however, AI agents need access to real user data to deliver personalized, context-aware experiences. This raises two major issues: 1. 𝐂𝐨𝐧𝐬𝐞𝐧𝐭 𝐂𝐨𝐦𝐩𝐥𝐞𝐱𝐢𝐭𝐲: Consent isn’t just a checkbox. It must be granular, auditable, and revocable. Yet today’s AI agents are often granted broad access — akin to full admin rights — to entire databases. This is highly risky and non-compliant. What if the AI agent shows the wife's transaction to the husband or worse to a stranger?  2. 𝐃𝐚𝐭𝐚 𝐀𝐫𝐜𝐡𝐢𝐭𝐞𝐜𝐭𝐮𝐫𝐞 𝐌𝐢𝐬𝐚𝐥𝐢𝐠𝐧𝐦𝐞𝐧𝐭: Most enterprise data is structured by function (e.g., payments, orders) rather than by user. So even if consent is obtained, we can't easily isolate and expose only a specific user's data to the AI agent. Instead, we give access to entire tables and rely on the agent to filter correctly — a fragile and error-prone approach I am sure that the first wave of commercial enterprise AI solutions will pay only lip service to user consent (if that). It will be argued within board rooms that AI tooling can be retro-fitted on existing legacy stacks and all concerns have been adequately handled. Its because everyone wants to get onto the AI bandwagon quickly and all risks are theoritical today It will only be after a few very public disasters that regulators will step in, real public debate on what is needed will happen etc. So this post is probably highly premature! But we've built Tachyon for the last decade on these very principles and are building Zeta's AI platform on the similar lines - User-centric data models that allow scoped access. - Consent-aware AI agents that operate within clearly defined boundaries. - Governance frameworks that enforce transparency, accountability, and fairness. Generative AI is not so much of a tech challenge as it is a design, ethics, and architecture challenge. Solving these will be key to unlocking its full potential in production.

Multi-agent AI systems can be exceptionally powerful. However they are very hard to configure. A new system leverages iterative feedback loops to automate the optimization process, resulting in significant performance improvements. The research paves the way to optimizing multi-agent workflows and agent interactions to improve system scalability, efficiency, and flexibility. Applications include healthcare, business process automation, AI-driven content generation, and many other industries. 🔄 Structured iteration for continuous improvement. The system operates through a self-improving cycle: the Hypothesis Generation Agent—powered by Llama 3.2-3B—proposes refinements, the Modification Agent implements changes, and the Execution Agent runs the updated system. The Evaluation Agent assesses outputs against qualitative and quantitative metrics, while the Selection Agent determines the best-performing configuration. This iterative loop continues until measurable performance gains plateau. 📊 Measurable gains through refinement. Case studies demonstrate that iterative self-optimization consistently improves output quality. The Market Research Agent saw a 0.9 improvement in clarity, actionability, and relevance, while the Career Transition Agent achieved 91% alignment with industry expertise. By continuously refining its structure, the system reduces output variability and enhances reliability, making its results more predictable and actionable. 🚀 Targeted modifications enhance agent performance. The Market Research Agent improved its strategic insights by introducing a Market Analyst and UX Specialist, while the Medical AI Architect Agent achieved a 0.9 regulatory compliance score by incorporating a Regulatory Compliance Specialist and Patient Advocate. The Lead Generation Agent, enhanced with a Business Development Specialist, increased data accuracy to 90%, improving lead qualification for AI-driven outreach. ⚙️ Role specialization drives efficiency and accuracy. Assigning domain-specific responsibilities—such as AI industry experts for pharmaceutical meeting facilitation and supply chain analysts for outreach campaigns—led to more precise insights, clearer recommendations, and higher engagement. Specialization ensures that each agent performs focused, high-value tasks, improving both the depth and relevance of AI-driven solutions. Refining and improving multi-agent systems and structures will be core to performance. I'll be sharing more as there are further developments in the space.