More than Parrots, Less than Gods

Glen Ansell
  • 0.0 (0)
  • Technology
  • Updated Weekly

Artificial intelligence isn't coming. It's already here — and most people are preparing for the wrong future. More Than Parrots, Less Than Gods is the no-hype guide to what AI actually means for your work, your business, and your life. Each episode breaks down one chapter from the book: how AI really works under the hood, which jobs are actually at risk, why 90% of corporate AI projects fail, and what skills will matter when the robots don't need us to hold their hands. This isn't tech evangelism. It's a research psychologist and 12-year AI practitioner looking at the data and he history.

  1. Jul 20

    Episode 19: The Next Five Years — A Map for Navigating the Next Five Years of AI Without Panic or Hype

    We have reached the end of a book about a technology still in its adolescence. That is a strange place to be — writing a conclusion about something that has not concluded, making predictions about a future that refuses to cooperate with prediction, and offering advice about a tool that will be different by the time you read these words. Welcome to Episode 19, the final episode. This is not prophecy. It is orientation — a map of the terrain ahead, drawn from the evidence we have and the patterns that have held steady through the noise. The next five years will not deliver the science-fiction future that evangelists promise, nor the apocalypse that doomers fear. It will deliver something more interesting, more complicated, and more human: a messy, uneven, contested transition in which winners and losers are determined not by the technology itself, but by the choices we make about how to use it. The most important shift will be cultural. Stanford HAI captured it: AI is moving from "Can AI do this?" to "How well, at what cost, and for whom?" The era of AI evangelism — every company needing an AI strategy, every product needing an AI feature — is ending. What replaces it is harder, slower, and less exciting: evaluation, integration, and incremental improvement. Most companies will report limited productivity gains outside targeted domains like programming and call centers. The companies that invested in data infrastructure and human expertise will pull ahead. The gap between AI performers and pretenders will widen. Agentic AI — systems that autonomously execute tasks across applications — will dominate the next five years. Gartner predicts 15 percent of work decisions will be autonomous by 2028. Yet Deloitte found that while 38 percent of organizations are piloting agents, only 11 percent are using them in production. The obstacles are not technical. They are organizational. Over 40 percent of agentic projects will fail because legacy systems cannot support them. Success requires clean data and clear objectives — not magical AI. The training data crisis will force a shift to synthetic data. High-quality text data may be exhausted. Synthetic data can fill gaps but also amplify errors and create feedback loops where models train on their own outputs until they become meaningless. Provenance will become increasingly opaque. AI sovereignty will fragment the internet. The UAE, South Korea, China, and the EU are all building domestic AI capabilities. Chinese models will operate under Chinese values. European models will emphasize privacy. American models will reflect commercial priorities. The universal chatbot is giving way to a patchwork of national systems. Robotics will move from demo to deployment in warehouses and factories. Brain-computer interfaces will restore function for the impaired, not enhance the healthy. Quantum computing will solve chemistry problems but will not transform AI training. The most likely scenario is not utopia or dystopia but the boring middle: gradual, uneven, contested change. The question is not whether AI will deliver heaven or hell. It is whether we can steer the boring middle toward outcomes that are marginally better rather than worse. What will remain for people to do? Not skills that machines cannot master, but choices that machines cannot make: what matters, what is worth protecting, and who is responsible when things go wrong. The question is not what AI will do to us. It is what we will do with it. The future is not determined. It is chosen. Choose wisely. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter #MoreThanParrots #ArtificialIntelligence #FutureOfAI #DigitalTransformation #AgenticAI #FutureOfWork #TechTrends

    Episode 19: The Next Five Years — A Map for Navigating the Next Five Years of AI Without Panic or Hype

  2. Jul 20

    Episode 18: Why Real AI Risks Are Boring - Why the Real Dangers Are Nothing Like the Movies

    When people imagine AI risks, they picture Hollywood: rogue robots, sentient machines, the intelligence explosion. Episode 18 explains why the real risks are exactly the opposite. They are boring, invisible, and already here. The most immediate risk is the collapse of shared reality. Deepfakes crossed a critical threshold in 2026. In Ireland's presidential election, a fake video showed the winner withdrawing days before polling. In the Netherlands, four hundred synthetic images attacked political candidates. The deeper problem is not that people will believe everything. It is that they will believe nothing. When any image or video can be faked, suspicion becomes universal. Democracy depends on a common factual foundation. AI erodes that foundation one pixel at a time. Job displacement is not mass unemployment. It is a two-track labour market where some workers are amplified and others are discarded. Research shows fifty to fifty-five percent of US jobs will be reshaped in the next few years, while ten to fifteen percent face elimination. The pattern is consistent. AI does not eliminate work. It eliminates the entry points. Junior developers find routine coding done by AI. Paralegals find standard contracts drafted by machines. The senior roles remain. The pipeline that feeds them is drying up. The most underreported risk is who controls the technology. Despite hype about new challengers, the dominant players remain the same five companies. They control the chips, the cloud, the models, and the talent. A technology that shapes what people see and believe is dangerous. When controlled by five companies with no democratic accountability, it is dangerous in a different way. Autonomous weapons raise legal problems that have no answer. International law requires human accountability for killing. A drone that mistakenly kills a child creates a responsibility vacuum. The programmer, the commander, the manufacturer, the machine: in practice, no one is accountable. Surveillance is ending anonymity in public spaces. You cannot opt out of walking past a camera. The state knows what you look like. You do not know what the state knows. The alignment problem — ensuring AI objectives match human values — is intellectually important but not the most urgent risk. The slow erosion of human agency is more worrying. When humans work with reliable AI, they gradually stop thinking for themselves. Each small surrender of judgment is rational. Over years, the accumulated surrender becomes a condition. The x-risk debate splits experts. Some warn of existential threat. Others dismiss it as doomerism. The reasonable position is to take both seriously and remember that boring risks kill more people than dramatic ones. The risks that actually matter are not the ones that make headlines. They are the ones that make headlines impossible to believe. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter 🎧 Subscribe for new episodes every week #MoreThanParrots #ArtificialIntelligence #AIRisks #Deepfakes #Misinformation #JobDisplacement #Surveillance #AIEthics #FutureOfWork #DigitalTrust

    Episode 18: Why Real AI Risks Are Boring - Why the Real Dangers Are Nothing Like the Movies

  3. Jul 20

    Episode 17: Why AGI Timelines Are Collapsing — What It Actually Means When Experts Say Machines Will Outthink Us

    In January 2025, Sam Altman announced: "We are now confident we know how to build AGI." Nvidia's CEO declared AGI had already arrived. Yet Ilya Sutskever — the scientist who built GPT-2, 3, and 4 — said the scaling era was over. How can the smartest people on earth disagree so completely? Welcome to Episode 17. This episode is about the most contested question in technology: when will machines become generally intelligent — and what does that even mean? The first problem is definition. AGI means one thing to a CEO raising money, another to a philosopher warning of extinction, and another to a chipmaker selling GPUs. To OpenAI it is "outperforming humans at most economically valuable work." To others it is human-level intelligence across all domains. When a term means everything, it means nothing. The second problem is scaling. For years, the dominant theory was simple: bigger models plus more data equals smarter AI. GPT-2 was a toy. GPT-3 was a curiosity. GPT-4 was a product. Each leap came from scale, not breakthroughs. But in 2025, the cracks showed. Reasoning models failed to generalize to messy real-world tasks. Most performance gains came from giving models more time to think — which is prohibitively expensive at scale. And reinforcement learning proved vastly less efficient than pre-training. The staircase of improvement continues, but each step is harder than the last. Then there is emergence — capabilities that appear suddenly as models grow. A model might fail at a math problem at 50 billion parameters and solve it flawlessly at 100 billion. Some call this evidence that intelligence will arrive unpredictably. Others argue it is a measurement illusion — coarse tests make small improvements look like magic. The truth is we cannot predict what will emerge next, which makes AGI timelines inherently uncertain. Expert predictions range from "already here" to "decades away." Altman says 2025 to 2028. Musk says 2025 to 2026. Hinton says 2028 to 2043. LeCun says 2029 to 2034. Sutskever now says 2030 to 2045. The short-timeline predictors have financial incentives to attract investment. The long-timeline skeptics have watched hype cycles crash before. No one has a crystal ball. Here is the nuanced truth. AI will keep improving, but the easy gains are over. The future is not one giant brain that does everything. It is specialized systems orchestrated by humans. Capabilities emerge, but within bounded domains. The real distinction is not AGI versus not AGI. It is impressive versus useful. A model that writes a novel is impressive. A model that reliably processes an insurance claim is useful. The boring risks are the real risks. Bias, overreliance, concentration of power, and erosion of human judgment. An AI that cannot think does not need to escape its data center to cause harm. It just needs humans who treat it as if it can think. AGI is either the most important or the most overhyped concept in AI. The future is not a cliff we fall off. It is a long, uneven staircase we climb with no guarantee the next step will be there. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter #MoreThanParrots #ArtificialIntelligence #AGI #FutureOfAI #TechPredictions #DigitalTransformation #AIResearch

    Episode 17: Why AGI Timelines Are Collapsing — What It Actually Means When Experts Say Machines Will Outthink Us

  4. Jul 20

    Episode 16: Why Smart People Trust Bad Algorithms — What Schools Won't Teach the Next Generation About AI

    In December 2025, a report revealed something quietly devastating: zero US states require students to complete any AI or computer science course to graduate high school. Meanwhile, 97 percent of high school and college students have already used AI for schoolwork. The students are not waiting for the curriculum. They are teaching themselves while the adults debate whether AI belongs in the classroom. Welcome to Episode 16. This episode is about preparing the next generation for a world the schools are not ready to teach. The essay — that centuries-old training ground for sustained thought — is dying. A New Yorker investigation found students using AI for every paper, earning A-minuses on work they barely understood. One student photographed museum wall text, fed it to Claude, and submitted the result. When asked if it was cheating, he replied: "Of course. Are you kidding me?" Faculty are retreating to blue-book exams and handwritten essays. It is a holding action, not a solution. The economy these students will enter does not reward handwriting speed. It rewards the ability to work with AI as a partner. The skill that matters is no longer writing from scratch. It is directing — formulating precise instructions, evaluating outputs, iterating, and integrating AI-generated material into a coherent whole. But the struggle of writing is where learning happens. Remove the struggle, and you remove the learning. Coding still matters, even though AI can code. GitHub Copilot now writes nearly half of production code. Yet developer productivity has barely moved. Why? Because coding was never primarily about typing syntax. It was about thinking — reading systems, debugging failures, designing architectures. AI handles the 20 percent of typing. It is unreliable at the 80 percent of thinking. Entry-level programming jobs have collapsed 67 to 73 percent since 2022. The market is not rejecting coders. It is rejecting routine coders while demanding systems thinkers. What should schools teach? Computational thinking. Data literacy. Verification skills. Algorithmic awareness. Ethical reasoning. Every student should know how to trace a claim to its source, recognize when a feed is manipulating them, and ask what data trained the model and who it might harm. Digital citizenship must expand beyond cyberbullying. Students need to understand that every technology has a business model — TikTok shows content to keep them scrolling, not to inform them. That data is power, and power is unevenly distributed. And that accountability cannot be automated. When an AI rejects a job application, someone built that system, someone profits from it, and someone must answer when it fails. Parents cannot wait for schools. Require unassisted work for core skills. Let children use AI for research, but insist that first drafts and math problems be done without assistance. Talk about what AI gets wrong. Model the behavior you want — use AI as a partner, not a replacement. And delay full access until they have developed the cognitive habits that AI erodes: sustained attention, independent reasoning, tolerance for ambiguity. The next generation will not be saved by a curriculum overhaul alone. They will be saved by millions of small decisions to preserve the difficult work of learning in a world that offers an easy alternative. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter #MoreThanParrots #ArtificialIntelligence #Education #FutureOfWork #DigitalCitizenship #GenZ #CriticalThinking #AIEthics

    Episode 16: Why Smart People Trust Bad Algorithms — What Schools Won't Teach the Next Generation About AI

  5. Jul 20

    Episode 15: Why Humans Stop Thinking Around AI — How to Rebuild Your Critical Judgment in an Age of Flawless-Sounding Machines

    In 2023, Stanford researchers ran an experiment. They took ninety-one essays written by non-native English speakers and fed them to AI detection tools. The result: the tools falsely classified 61 percent of human essays as AI-generated. Nearly all essays were flagged by at least one detector. The detectors were not catching cheaters. They were punishing students for writing simply in a second language. Welcome to Episode 15. This episode is about critical thinking in an age designed to outsource it. The AI detection arms race is the wrong battle. Even a perfect detector would flag falsehoods only after they were read and believed. The real threat is not AI-generated content. It is truthiness — the property of sounding fluent, confident, and authoritative without regard for accuracy. AI packages falsehoods with perfect credibility markers: citations, statistics, and measured recommendations. Often the citations are fabricated, the experts never spoke, and the statistics are invented. But the form is perfect, so we believe it. Then there is bias. AI does not create bias. It inherits and amplifies it. The COMPAS recidivism algorithm was twice as likely to misclassify Black defendants as high risk. A 2025 Stanford study of 3.4 million job applicants found that 26 percent of Black applicants and 15 percent of Asian applicants faced AI discrimination. Because 90 percent of employers use the same few vendors, a single flawed algorithm can shut qualified candidates out of entire industries. And when Google's image recognition mislabeled a Black couple as gorillas, the fix was not to improve the algorithm. It was to remove gorillas from the labeling system entirely. The deeper problem is human. Research on AI complacency found that employees intentionally stop checking AI output even when they spot factual errors. The driver is not laziness. It is diffusion of responsibility — when humans supervise machines, neither feels fully accountable. The result is authority without accountability. A hiring manager rejects a qualified candidate and points to the AI score. A lender denies a mortgage and cites the risk model. The human retains formal authority but delegates moral responsibility to a system that cannot be questioned. So what do we do? Five mental models. Distrust fluency. If you would not believe a poorly written version of the same claim, you are being swayed by style, not substance. Trace the source chain. For every significant claim, find the original source. Does it say what the AI claims? This takes time. It is the only reliable defense. Demand accountability. Before accepting any AI-assisted decision, ask who is responsible if it is wrong. If the answer is unclear, the decision should not stand. Assume bias. Models are trained on historical data, and history is full of prejudice. Every recommendation should be filtered through the question: what groups might this harm? Add friction. The systems that make AI easiest to use often make it hardest to verify. Ask for a second opinion. Write your own answer first. Take notes by hand. The goal is not to reject AI. It is to maintain the habit of independent judgment. The more powerful the tool, the more important the judgment of the person using it. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter #MoreThanParrots #ArtificialIntelligence #CriticalThinking #AIBias #DigitalTransformation #FutureOfWork #AIEthics #Misinformation

    Episode 15: Why Humans Stop Thinking Around AI — How to Rebuild Your Critical Judgment in an Age of Flawless-Sounding Machines

  6. Jul 20

    Episode 14: Mastering the Jagged Frontier of AI — Your Practical Toolkit for Working With Machines Without Losing Your Mind

    In 2024, Microsoft found that 75 percent of knowledge workers now use AI at work. Nearly half had started in just the previous six months. AI became a workplace staple faster than email or smartphones. But here is what did not make the headlines: most of them are using it badly. Welcome to Episode 14. This episode is your practical toolkit for working with AI as a competent partner rather than a crutch. First, build a deliberate stack. Do not collect AI tools like Pokemon cards. Pick one foundation model and know it deeply. Claude excels at long documents and nuanced writing. ChatGPT has the largest ecosystem. Gemini integrates with Google Workspace. Perplexity searches the live web and cites sources. Choose one that fits your workflow. Second, layer your tools. Use AI for email triage and meeting transcription to cut the mechanical 70 percent of communication. Use creation tools like Copilot for code or Jasper for marketing copy only if your role demands it. Use automation platforms like Zapier to connect software and eliminate repetitive steps. Use knowledge management tools like Notion to make your notes searchable and queryable. Third, prompt clearly. Good prompting is not wizardry. It is clear thinking written down. Define the role. Provide context. Specify the format. Iterate in conversation. Use constraints creatively. The models in 2026 are dramatically better than 2023 at understanding vague requests, but intentionality still determines output quality. Fourth, verify everything. AI does not know what is true. It knows what is statistically likely. Top models hallucinate sources at rates between 0.1 and 0.7 percent. That sounds low until you interact with AI daily and encounter a fabricated citation roughly once per month. Stanford found that one in three citations in AI legal research tools were fake. Never publish AI output without checking the facts. Fifth, know the cost. A typical worker subscribes to five to eight AI tools at twenty to forty dollars each per month. That is one thousand to four thousand dollars annually. If you spend more time learning tools than saving time with them, your stack is too big. Most importantly, protect your mind. Research shows that doctors who used AI to detect cancer eventually became worse at identifying it independently. MIT found that AI-assisted fake news detection improved immediate accuracy by 21 percent, but by week four unassisted performance had degraded 15 percent below baseline. Heavier AI users score worse on critical thinking tests. AI-assisted creative work is more predictable and less original. Use the blank page rule. Fill the page yourself before opening AI. Write your rough ideas first. Let your brain make unexpected connections before consulting the machine. The first draft should be human. The polish can be AI. The winners will not be those with the most subscriptions or fanciest prompts. They will be those who treat AI as a tool to amplify human capability, not replace it. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter #MoreThanParrots #ArtificialIntelligence #AIProductivity #FutureOfWork #CriticalThinking #DigitalTransformation #AITools #CareerAdvice

    Episode 14: Mastering the Jagged Frontier of AI — Your Practical Toolkit for Working With Machines Without Losing Your Mind

  7. Jul 20

    Episode 13: Why Ninety Percent of AI Projects Fail — What the 5% Who Succeed Do Differently

    In 2025, organisations spent an estimated one and a half trillion dollars on artificial intelligence. Roughly ninety percent of those projects failed to deliver measurable value. This is not a technology problem. It is a strategy problem. Episode 13 explains why so many AI initiatives collapse and what the small minority who succeed do instead. The single most common mistake is starting with the technology and looking for problems to solve. A bank launches an AI initiative because competitors have one. Eighteen months later, they have a chatbot that answers twelve percent of queries correctly and annoys the rest. The same bank, starting with the pain point — mortgage approvals taking eleven days with forty percent abandonment — might use AI to automate document verification, cut the timeline to two days, and recover millions in lost revenue. The technology is identical. The framing is everything. The chapter introduces three questions every leader must answer before spending a dollar. What routine work consumes your smartest people? Where does speed matter more than creativity? What decisions need better data? These questions force discipline before technology selection. The buy versus build versus partner decision matters enormously. Internal builds succeed roughly one-third as often as vendor partnerships. Buying works for commoditised use cases. Building makes sense only when the use case is so specific or sensitive that external access is impossible. Partnerships combine external expertise with internal domain knowledge, succeeding about two-thirds of the time. Measuring return on investment is where most projects die. A proof of concept showing ninety-four percent accuracy in a controlled environment is not evidence of business value. The real questions involve messy real-world data, costs at scale, actual resolution rates without human intervention, and customer satisfaction impact. These answers come only in production, often too late. Change management is the hardest part. Research shows AI value breaks down as ten percent algorithms, twenty percent technology, and seventy percent people and workforce transformation. Organisations that invert this ratio consistently fail. The five percent who succeed invest heavily in upskilling, with managers actively role-modelling AI use daily. They embed AI into existing workflows rather than forcing new tools and processes. They address employee fear directly, communicating clearly about which tasks change and which roles evolve. Data quality is the most underestimated foundation. Eighty-five percent of AI projects fail due to poor data quality or lack of relevant data. Only twelve percent of organisations report data sufficient for AI. Most businesses have spent decades accumulating data in formats optimised for human reports rather than machine consumption. Converting this legacy is a multi-year investment that must begin before any models are trained. The episode also covers when AI is simply the wrong tool. Poorly defined problems, missing data, high-stakes decisions requiring human judgment, rapidly changing processes, and misaligned stakeholders are all reasons to say no. The discipline to decline is as important as the vision to pursue. If you are a business leader wondering how to make AI matter for your organisation without joining the ninety percent who waste their money, this episode provides the strategic framework built on the wreckage of a trillion dollars in failed projects. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter 🎧 Subscribe for new episodes every week #MoreThanParrots #ArtificialIntelligence #BusinessStrategy #AIProjects #ROI #ChangeManagement #DataQuality #MachineLearning #FutureOfWork #TechLeadership

    Episode 13: Why Ninety Percent of AI Projects Fail — What the 5% Who Succeed Do Differently

  8. Jul 20

    Episode 12: How Weaponized AI Shatters Human Trust - The Same Technology That Heals Can Also Harm

    Every AI capability that helps humanity can also be turned against it. Episode 12 explores the dark side of artificial intelligence — not as science fiction, but as documented reality in 2026. The same neural network that spots cancer in a medical scan can generate a fake video of a politician taking a bribe. The same algorithm that optimises your supply chain can manipulate a stock market. The same system that tutors a struggling student can craft a phishing email convincing enough to trick a seasoned executive into wiring millions to criminals. Cybersecurity has spent decades defending against predictable threats. Antivirus software scans for known signatures. Firewalls block recognised patterns. The entire defence assumes attackers use fixed toolkits. AI breaks that assumption. In 2026, researchers demonstrated an AI-driven computer worm that discovers and exploits vulnerabilities in real time, propagating through networks without human intervention. It does not need a list of pre-loaded exploits. It writes its own. The internet of things makes this worse. Sixteen billion connected devices — cameras, thermostats, medical monitors — run outdated software with default passwords. A compromised smart camera in a suburban home becomes a launchpad for attacking the neighbour's router, which bridges into a corporate network. The AI does not need to trick a human. It simply scans, exploits, and spreads. Disinformation has evolved from clumsy propaganda to precision-targeted cognitive warfare. AI bots no longer broadcast obvious spam. They join conversations, express sympathy, share links, ask questions, and gradually nudge opinions. One human operator can manage hundreds of fake personas, each with its own writing style and fake history. Deepfakes have moved from novelty to mainstream criminal tool. A manager in Hong Kong authorised a thirty-five million dollar transfer after a video call with what appeared to be the company's chief financial officer. Every person on the call was a fake. AI-generated phishing emails achieve click-through rates four times higher than human-written attacks because they eliminate the awkward grammar and suspicious formatting that used to give scams away. The surveillance applications are equally consequential. Facial recognition has caused wrongful arrests in the United States, with higher error rates for people with darker skin. Predictive policing algorithms trained on biased historical data send more police to already over-policed neighbourhoods, creating feedback loops of discrimination. China's social credit system combines facial recognition, digital tracking, and behavioural analysis to evaluate and punish citizens in real time. The structural problem is asymmetry. Attack is easier than defence. A deepfake requires one person with a laptop to create. Detecting it requires building and deploying detection models across every platform where video is shared. A phishing email takes minutes to write. Defending against it requires retraining millions of employees and updating filters constantly. This episode matters because the dark side is not a niche concern. It is the water we all swim in. When banks add verification layers to fight fraud, the cost passes to consumers. When political disinformation poisons democratic discourse, the resulting paralysis affects healthcare, education, and infrastructure. When surveillance normalises the idea that privacy is an anachronism, freedom narrows for everyone. The question is not whether these risks exist. They do. The question is whether we have the institutions, the wisdom, and the will to manage them before they manage us. 📖 Get the book: https://morethanparrots.com 📧 Free chapter slides: https://morethanparrots.com/newsletter 🎧 Subscribe for new episodes every week #MoreThanParrots #ArtificialIntelligence #AI #CyberSecurity #Deepfake #Disinformation #Surveillance #FutureOfWork #Privacy #DigitalTrust

    Episode 12: How Weaponized AI Shatters Human Trust - The Same Technology That Heals Can Also Harm
See All (18)

About

Artificial intelligence isn't coming. It's already here — and most people are preparing for the wrong future. More Than Parrots, Less Than Gods is the no-hype guide to what AI actually means for your work, your business, and your life. Each episode breaks down one chapter from the book: how AI really works under the hood, which jobs are actually at risk, why 90% of corporate AI projects fail, and what skills will matter when the robots don't need us to hold their hands. This isn't tech evangelism. It's a research psychologist and 12-year AI practitioner looking at the data and he history.

Information