You said:
Summarize the transcript in hierarchal bullet points. summarize with all interesting and important notes and some obscure and/or abstract facts. Start with the key takeaways. Conclude with a paragraph explaining the overall theme in a simplified essay
Title: "Why AI Is Tech's Latest Hoax"
Transcript: "Tech is a sector unlike any other - it’s an industry where individuals can turn into billionaires overnight, technical ideas supersede business fundamentals, and leaders are rewarded for showmanship over competence. In today’s Silicon Valley, innovation is crowned, not earned to those who can tell a story and look the part of an eccentric genius. Venture capitalists and founders are symbiotic - VCs need radical ideas and founders want to start businesses on someone else’s dime. Un profitable companies are kept alive with injections of capital, gamed valuations, and manufactured media hype with the goal of surviving long enough to IPO or be acquired. The inspirational speeches, motivational essays, generous salaries, flashy perks, and founder features are no accident. It’s all there by design to keep the talent pipeline of new grads flowing to the sector. Founders lean on VCs for access to capital, talent and operational mentorship while VCs lean on founder s to educate them on technology. But VCs aren’t prophets. Most are former Wall Street bankers or celebrities who have never worked in tech. It’s the blind leading the blind and both sides must perform to reach the same payday. When you win in tech, you win big - hence, you only need to win once. This is why there is a near-infinite pool of aspiring founders and VCs and why both parties are so quick to forgive, reconcile, and work together on “the next big thing”. Yet whenever inve stors turn bearish on tech, it doesn’t take long for the sector to come up with a new growth story. The most recent was in 2022 when the public markets soured on big data and SaaS. Starting in the early 2010s, Silicon Valley had championed big data as a revolutionary technology that could unearth deep insights, hidden patterns, and innovation from massive amounts of never-before-captured data. Big data promised a new sophisticated data-driven world where one could precisely predic t demand before it even existed, trends before they started, and behavior before it occurred - and there was immense potential for the public and private sector. Police could prevent crime before it happened, researchers could detect cancer before it spread, and companies could optimize products, make correct decisions every time, and gain a powerful edge. Big data was a movement as much as a technology and SaaS had been positioned as a resilient, high-margin, low-churn, profita ble business model for modern software companies. Yet the market started to question if any of these promises had been real as the vast majority of consumer startups and SaaS companies were still bleeding years after IPO. Even in the most favorable, low-interest business conditions in history, there were barely any winners to point to. Out of nowhere, ChatGPT was released and AI became Silicon Valley’s next big thing. In less than 2 years, everyone has forgotten about big data a nd SaaS. Every tech company is now an “AI company”, every Fortune 500 needs an “AI strategy”, VCs are only investing in AI startups, everyone’s title on LinkedIn mentions AI, and every product is an “AI” product. To maintain hype, AI was brought to the public sector. Silicon Valley figureheads put on a performance in front of Congress, begging for regulation, urging protection for workers whose jobs would be displaced, and fear mongering about a future apocalypse. This song-and-d ance was done over and over again until the White House was spooked but more importantly, the public was convinced of the immense potential of AI. And people continue to speculate with conviction that artists, animators, translators, and programmers are all next in line to lose their jobs. This episode is not a technical debate, but rather a deep dive into how AI is just the latest tale spun by Silicon Valley to sweep the prior failed trends under the rug, keep valuations high, an d the outlook positive. Before AI, there was crypto, web3, blockchain, virtual reality, augmented reality, big data, IoT, and wearables - all supposedly revolutionary technologies that have not lived up to the hype. In this episode, we’ll dive into the real market dynamics that push companies and individuals to jump headfirst into these tech trends, how this all started with big data, and why AI is ultimately just another pump-and-dump. This episode is sponsored by Kajabi, the ultima te all-in-one platform that helps entrepreneurs build successful online businesses by unlocking predictable, recurring revenue. As a creator, it’s difficult to build a brand as just about every platform, tool, or partner out there that promises to help you asks for a percentage of your profits. As a creator, you deserve all the benefits, upside, and all of the revenue you earn. At the same time, it takes lots of time, effort, and unavoidable trial-and-error to build a great tea m, attract talent, and establish a positive culture for your company. Most creators end up doing it all or trying to do it all themselves which ultimately only limits growth and encourages unsustainability. With Kajabi, you can build your business the way you want and keep every dollar that you earn. And Kajabi gives you all the tools you need to build a profitable business, so you don’t have to hire a huge team or do it all alone. No matter your niche, Kajabi makes it easy to tur n your skills, passions, and experiences into enriching online courses, exclusive membership sites, subscription podcasts, thriving communities, personalized coaching, and more. The best part? Kajabi doesn’t take a cut of your revenue, because everything is owned and controlled by you. So you keep 100% of what you earn! And with Kajabi you also get robust analytics, easy payment options, email marketing tools, and customizable website templates — all built in. You don’t need a hug e audience to make sustainable income. There are thousands of creators on Kajabi making six and seven figures with less than fifty thousand followers! Right now, Kajabi is offering a free 30-day free trial to start your business if you go to Kajabi dot com slash MODERNMBA. That’s K A J A B I dot com slash MODERNMBA. Kajabi dot com slash MODERNMBA and join the creators and entrepreneurs who have made over six billion dollars! The late 2000s was a period of genuine innovation - the introduction of smartphones and tablets had created new markets. It was the advent of the mobile Internet, geolocation, 3G, and the app store. More could be accomplished in this online world than ever before. You could now track the location of users, which was data that had not been available before. And if you designed for mobile, users would organically flock to you for that superior experience. Numerous startups emerged, eager to be the first movers on these new platforms. On e was Groupon, an online platform that sent out coupons for local businesses. Their pitch was that they were amassing unprecedented data on consumers and driving paying customers in the door for merchants - which was more than what Google had to offer. They knew where people lived, their age, interests, behavior and fed all that data into algorithms to send only the most relevant, enticing coupons. And as the volume of coupons and merchants increased, the more data collected, the faster Groupon could push customers to a business, and the more money it could make. Pandora’s story was identical to Groupon’s - they scientifically abstract every song down to 480 technical attributes, collect data on each listener to figure out what they like, use proprietary algorithms to predict what music they’ll like next, and generate personalized playlists. The more listeners, the more data, and the more accurate its algorithms would get at playing the right song at the right time for each listener. Since they knew the age, gender, and location of every user, they could sell to advertisers. Another popular app was Yelp who asserted user data and network effects as their moat. Because Yelp knew who was looking for what and where, like Pandora, they sold these eyeballs to advertisers. GrubHub went in a different direction and focused on the takeout experience. Restaurants didn’t need to build apps - instead, they could pay GrubHub to support online a nd mobile ordering overnight. And GrubHub had algorithms to personalize feeds and predict what users wanted to eat when. Other first-movers in this period were Etsy, Foursquare, Twitter, and WhatsApp. By the early 2010s, the narrative had broadened. Smartphones went mainstream, apps became commonplace, and mobile experiences became the norm. It was at this time when Marc Andressen published his famous essay “Why Software is Eating the World”. The thesis was simple - data was go ld in this new digital world. Data could be applied to unlock business value, product optimization, and personalization in any industry. Users generated data, data powered algorithms, and algorithms produced innovation. Zynga, the studio behind Farmville, proclaimed that they were innovators because of data. They looked at monetization, retention, engagement every day to quantify what got people hooked and kept whales spending. As Zygna amassed more users and data, they alleged th eir analytics would only become more precise and profitable. As an online furniture store, Wayfair offered a near-infinite amount of furniture, styles, and brands. They talked up their real-time proprietary personalization, inventory, and pricing algorithms that would maximize profits on every piece of furniture. Chegg looked to do the same in education with a social network that collected data on high school students. While that product never succeeded, even Chegg’s leaders unde rstood the narrative. “There is a playbook, followed by many companies, such as Facebook, LinkedIn, Netflix, or Spotify, which is to take a giant category, build products and services that consumers value, use technology to deliver them at scale and leverage the data it generates. In the case of Facebook, it’s social, LinkedIn, professional, Netflix, entertainment, and Spotify, music. The advantages these companies have built through their data makes each of them very hard to compet e against. If you own your customer, the channel of distribution, you collect data, you own that data, and you’re able to use that data to improve your product, to monitor what people do to deliver better products and services, you should be in the best position to provide overwhelming value to your customer base and to build a giant moat.” Thousands of new consumer startups flooded the market backed by venture capital and singing the same narrative that data is power. But these startups did not have the organic adoption and network effects that the first-movers like Yelp, Pandora, LinkedIn, and Facebook had. As a result, they needed to spend millions of dollars on advertising to acquire users. This chase for users and their data was used to justify the unsustainable marketing spend and broken unit economics. The spin at the time was that these startups were all unprofitable by design as they needed to acquire the necessary data to supercharge their produ ct and business. There was Wish, who boasted that its algorithms could get customers hooked on buying $1 junk online and OpenDoor who bragged that they had so much data that their algorithms could value real estate better than humans and flip properties anywhere for profit. Casper proclaimed that data powered their disruption of the mattress industry and DoorDash asserted that it was data that enabled fast delivery, order routing, and strong earnings for gig workers. And then there was Affirm who contended that they had the data to issue micro-loans for luxury purchases. These are just the startups that we’ve covered in the past on Modern MBA. There were many more like Blue Apron, who claimed to have so much data that they knew every individual’s taste profiles. They could algorithmically predict demand for any given meal kits and achieve production efficiencies no one else could. Warby Parker pushed that they had data that no other optical company possessed and were so data-driven that every pair of glasses would be a best-seller. AllBirds pounded that they knew more about customers than traditional sneaker brands and over time would have higher margins and as many loyal customers as Nike. Stitch Fix declared that they had more data on customers than any department store or fashion brand and its algorithms could accurately predict what clothes someone would buy sight-unseen. TrueCar promised to transform car-buying with algorithms that ingested every sale in history and could generate the most accurate price for any vehicle with the make, model, and year alone. Lemonade billed itself as a disruptor whose algorithms crunched so much data around the clock that they could process claims and underwrite policies faster and cheaper than traditional insurance companies. LendingClub crowed that their data unlocked lending efficiencies and in turn allowed them to algorithmically issue lower-interest loans online wit hin minutes. SoFi trumpeted that they were collecting all the behavioral data on customers that conventional banks did not, which would translate to better accuracy and greater profits as a lender. FitBit disrupted health and fitness by quantifying and visualizing every individual’s physical activity. Roku and Netflix both asserted that they had so much data as streamers that they understood viewers better than any film studio or cable network. They knew what people wanted to watc h and by extension, had the secret sauce to make any show or movie a hit. These are just the few out of the thousands of consumer startups that emerged in this period. If you look at the S-1 of any consumer tech IPO of the past 20 years and search for the word “data”, you will see this same narrative spelled out. It was a convincing story. In what world could data not be useful? Yet tech, contrary to what Silicon Valley advertises, is not a magical frontier where everyone can be a winner. The reality is that in every industry, there can only be a few winners. Yet no one could argue against the prosperity of Facebook, LinkedIn, Amazon, Google, Netflix, and Microsoft, who were all making a killing with data. Their earnings were snowballing in ways that the public markets had never seen before. The only thing more impressive than revenue was margins, which were the envy of the private sector. There was no other stock as reliable, valuable, and high-potential as FAANG. If data was the moat for these tech leaders, why couldn’t it work for a startup that was nimbler, faster, and concentrated? If a startup could apply the same playbook to a smaller industry and achieve just 1/10th of what Facebook or Netflix had accomplished, it would still be enough to IPO. With such high-flying results and thousands of VCs and startups all crowing the same story - data became fashionable. Every one of these consumer startups were crowned as innovators ba sed on user growth alone, which was misleading given how much of it had been attained through heavy advertising and artificial subsidies. Data was billed as the means to unlock innovation. But beyond selling data to advertisers, no real business value had been discovered. But by the mid-2010s, many consumer startups were starting to flame out. The few that had gone public were losing just as much money, if not more, than they were at IPO years before. The premise of innovation t hrough data seemed less convincing by the quarter. In response, Silicon Valley moved the goalposts once again. Data was still valuable - you just didn’t have enough of it or the means to interpret it. Basic analytics and personalization was no longer enough. What you needed was terabytes of data, sophisticated tools, and data scientists to get to the promised land. This new trend was called Big Data. To maintain valuations and reputations, these consumer startups embraced Silicon V alley’s latest narrative. Still, the Fortune 500 companies were spooked. No CEO wanted to be caught with their pants down, no executive dared to say that their teams were not data-driven, and every Wall Street analyst wanted to know how they were going to stop tech from eating their lunch. The conversation quickly devolved into a pissing match of who had the most data and best culture. Groupon’s newest CEO went all-in. “Groupon is in the data stream for every business transaction. We see every bit that comes through these businesses which gives us really critical insights. We’re rewriting our basic personalization and relevance with advanced techniques and machine learning. The secret to our methodology is a data-driven approach. We have more than 9 petabytes of data and we A/B test every single feature. How many companies do this? Not a lot.” None of this would stop Groupon from bottoming out in just 5 years. Zynga - “We have a really strong team of data scientists who look at the relationship between the ads we’re serving, the number, type, unit, and impact on player engagement. We have developed specialized algorithms and machine learning. We’re never going to have the same user data as Facebook but we can get close. That’s how we stand out when we’re talking to advertisers with data-science focused engagement.” Revenue improved yet Zynga remained as unsustainable as ever. Wayfair - “We capture 4 terabytes of data every day and 40-billion customer actions a year. We have a depth of data rare within the home category. If you don’t have the ability to take advantage and manipulate the data for deep analysis, it’s tough. Over the last 4 years, we have built a team of 1,900 engineers and data scientists. Data science and machine learning influences our personalization, dynamic pricing, algorithmic merchandising, demand forecasting and advertising. As a result, we have been able to build multiple platforms at a strong ROI. Deep analysis of data is central to our business and how we win with customers.” 6 years later, Wayfair has not stemmed the bleeding and continues to lose nearly a billion dollars every year. Blue Apron - “Our direct-to-consumer platform is our most valuable asset which provides extensive behavioral insights to drive innovation. We have touched millions of customers and have a lot of data from our 6 year history. We use machine learning to give a sense of what custome rs are likely to order. This allows us to order the proper amount of protein and produce, which directly impacts food costs and margins.” None of this helped pull Blue Apron out of its downward spiral. GrubHub: “We have data on over 100 million orders. Our algorithms will get smarter about the most popular dishes in every neighborhood in 900+ cities. There is no other company in the U.S. that has this level of transactional data. The velocity of high-quality, high-fidelity data th at we’re aggregating is incredible at 70,000 points per day. The learnings from our massive troves of data and data-driven insights will position us well for years to come.” GrubHub’s losses have deepened and its owners are still struggling to find anyone willing to take the company off their hands. LendingClub: “We’ve issued over $40B personal loans in 10 years so the data we have generated is really massive. That is a big data advantage. We simply have this big scale that allows us to slice and dice customer profiles, create unique experiences, and underwriting processes. We deploy the latest machine learning to derive more than 100 customized and behavioral attributes, half of which are proprietary and based on our unique data assets.” LendingClub has since cratered in valuation, cut back its lending business, and is worth less than its revenue. Pandora: “Our biggest strength is the wealth of data and data science capabilities. We have built our product nd at much lower cost, which provides us more access to data and allows us to provide a better service to members. ” Sofi’s valuation has since plummeted as the company’s “innovation” is just selling the same lending products that banks have done for generations albeit with nicer UI. And Stitch Fix: “Our whole business model is predicated on this amazing data that we have. Fundamentally, data helps us buy more of the right product and get it into more of the right people. We have hms to cut costs - and still, has only continued to bleed money and users year after year. And we know how things have turned out for OpenDoor, Affirm, Wish, Casper, and DoorDash. It took until the early 2020s for the big data narrative to die out. No consumer startup had demonstrated anything meaningful with their big data and no one was going to wait another 10 years for progress. Yet even as the walls closed in, these startups held on, screaming about machine learning, deep learn AT&T, Comcast, Disney, Pepsico, Chase, Citigroup and other big companies each committed hundreds of millions of dollars - some, even billions, to build out such capabilities for themselves. They were all eager to signal that they could be just as cutting-edge as tech startups and publicly flaunted their investments in big data. Yet their adoption was fueled by fear rather than merit. No CEO wants to be the one who screwed the pooch. It was safer to have an iron in the fire than to talent skyrocketed. These were greenfield technologies and it was believed the more bodies you could throw at complexity, the faster you would arrive at a solution. Job opportunities and salaries for statisticians, data scientists, and software engineers reached record highs as the private sector competed for technical talent. But this only addressed the problem of “who” and not “how”. The majority of consumer startups and big corporations lacked the technical means to extract, s wed year after year towards these B2B startups. Unlike the VC-backed consumer startups, the corporations had the appetite and runway to spend 7-8 figures in perpetuity on any vendor who could help them get to the promised land. If we revisit the tech IPOs of the past decade and look at the companies with the greatest appreciation in valuation since IPO, the winners are nearly all enterprise startups. DataDog and Splunk are two leaders in telemetry that to this day have reached bil demand for data and engagement meant greater emphasis on SMS and email - which Sendgrid, RingCentral, and Twilio were happy to provide for a price. The volume of information, sensitivity of the data, and complexity of application logic made these companies targets for hackers. Security vendors like Cloudstrike, Barracuda, Okta, and Palo Alto Networks have all thrived under the promise of helping enterprises and startups secure their data. And to accelerate execution, startups lik ined at a big company, it’s difficult to rip out. Yet not every Fortune 500 had the talent to execute and most needed to pay an outside firm to perform the implementation. The money flowed not just to tooling, but also to consulting. Accenture, HP, IBM, Oracle, Booz Allen Hamilton, and Gartner all pledged that they had the expertise to pull off any big data project. It’s no surprise that these same consultants are now tooting their horns about AI. Ultimately, the companies that pr people are using the cloud and whatever trending technology drives them to use more of it, Amazon, Google, and Azure all win. Software engineers have always been in high demand but the paradigm shifted between the 2010s-2020s with big data. Across the private sector, companies were on the hunt for practitioners offering the highest salaries to the few with real-world experience. Since the technology was so new, there were no best practices. Everyone was figuring it out as they w maximizing compensation and hireability. Developers today are no longer coding grunts but instead vocal rock stars who can make demands and spearhead change. They get to decide on the behalf of their companies what clouds to use, what tools to adopt, and what vendors to partner with. This is why enterprise companies and cloud vendors spend millions every year on conventions, pizza, and beer just to court developers and convert them into champions of their products. It’s like a batt nd their VPs can justify budgets, get raises, and climb the ladder. The people that worked on big data had a vested interest in keeping the technology trendy despite the lack of business results - and are now doing the same with AI. Whether it’s bottoms-up or top-down, money talks - even for those who are simply adopting the technology itself. That’s why there’s never any end to the online debates between React vs Angular, Golang vs Rust, Kubernetes vs ECS, and so on. You can see th wn. But the current premise of AI is even more confusing - the new narrative is that data is inherently too complex. Instead, what we’re supposed to believe is that all that promised innovation, buried insights, hidden business value is still in this data - it’s just that humans can’t pull it out. Instead, we should trust this artificial being, where only a few have an understanding of what is really happening under the hood, for perspective and insights. We should be told the ans s is how to execute” and “this is how you can replicate step by step what we did for yourself”. Engineers essentially are re-inventing the wheel but never finishing. Managers and executives are pushing for vanity updates and exaggerated progress for the sake of growing their organizational influence, appeasing their higher ups, and securing their own promotions. The same is happening and will happen with AI. The market dynamics that drove big data are now playing out once again wi ber, ChatGPT was built on the same playbook of skirting regulation. ChatGPT went viral for its breadth of knowledge and convincing human-like prose - but when all of it is built on stolen data - who holds the power, the people who generated the data or the technology that summarized it? Like every other Silicon Valley technology of the past decade, AI is just another cash grab. The reason why OpenAI’s employees revolted when Sam Altman was fired was because his removal made their not make any difference for the consumer startups and Fortune 500 over the past decade - if anything, it prolonged the lifespan of the startups that should have never existed in the first place. These days, all companies are doing layoffs, cutting costs, engaging in shrinkflation, and buying back shares to squeeze profits just like they’ve always done - things that they shouldn’t have to do if big data actually generated the innovation and greater business efficiencies that was pro their shares upo"
