Despite its hefty $1.3 billion investment, the recent collapse of Inflection serves as a stark reminder of the volatile AI startup landscape. Inflection’s flagship product, Pi, a ChatGPT rival, failed to gain traction, leading to the company’s dismantling by Microsoft. This case exemplifies the broader trend of massive capital influx into AI ventures lacking substantial products.
The Rise and Fall of Inflection
Inflection was founded by notable entrepreneurs such as Mustafa Suleyman of DeepMind, Karén Simonyan, and Reid Hoffman. Suleyman, a co-founder of DeepMind, had previously contributed to its advancements in AI, which eventually led to its acquisition by Google. Simonyan brought extensive experience from his work on AI research, while Hoffman, co-founder of LinkedIn, provided substantial entrepreneurial and investment acumen.
With backing from influential investors including Bill Gates and Eric Schmidt, Inflection aimed to create a more empathetic AI companion. The company took around two years to develop Pi, its primary product, hoping to leverage its founders’ reputations and the significant capital raised to break into the AI market.
Why Pi Failed
Pi’s failure is attributed to several factors:
Lack of Unique Value: Pi’s context window was significantly shorter than competitors, hindering its ability to provide sustained conversational quality.
Market Oversaturation: The AI companion market is fiercely competitive, with established players like ChatGPT and Character.ai leading the pack.
Financial Mismanagement: Heavy investment without a corresponding viable product highlighted the risks of capital-heavy ventures in AI.
AI Funding and Startup Failures
The AI sector saw an estimated $50 billion in investments in 2023 alone. However, many startups have failed to deliver on their promises. Some notable closures in the last 18 months include:
Inflection: Absorbed by Microsoft, ceasing independent operations.
Vicarious: Acquired by Alphabet, failing to achieve its goal of human-like AI.
Element AI: Acquired by ServiceNow after struggling to commercialize its research.
Startup
Total Investment ($M)
Years to Product Launch
Peak Annual Revenue ($M)
Outcome
Inflection
1300
2
5
Acquired by Microsoft
Vicarious
150
4
2
Acquired by Alphabet
Element AI
257
3
10
Acquired by ServiceNow
MetaMind
45
2
1
Acquired by Salesforce
Geometric Intelligence
60
1
0.5
Acquired by Uber
The Future of AI Investment
This trend of high investment but low product viability raises concerns about the future of AI innovation. Consolidation around major players like Microsoft, Google, and OpenAI could stifle competition and limit diversity in AI development.
Conclusion
The downfall of Inflection underscores the precarious nature of AI investments. As the industry continues to grow, investors must prioritize viable, innovative products over mere potential. This shift could foster a more sustainable and dynamic AI ecosystem.
Is the AI Boom Overhyped? A Look at Potential Challenges
The rapid development of Artificial Intelligence (AI) has fueled excitement and hyper-investment. However, concerns are emerging about inflated expectations, not just the business outcomes, but also from the revenue side of the things.. This article explores potential challenges that could hinder widespread AI adoption and slow down the current boom.
The AI Hype:
AI has made significant strides, but some experts believe we might be overestimating its near-future capabilities. The recent surge in AI stock prices, particularly Nvidia’s, reflects this optimism. Today, it’s the third-most-valuable company globally, with an 80% share in AI chips—processors central to the largest and fastest value creation in history, amounting to $8 trillion. Since OpenAI released ChatGPT in October 2022, Nvidia’s value has surged by $2 trillion, equivalent to Amazon’s total worth. This week, Nvidia reported stellar quarterly earnings, with its core business—selling chips to data centres—up 427% year-over-year.
Bubble Talk:
History teaches us that bubbles form when unrealistic expectations drive prices far beyond a company or a sector’s true value. The “greater fool theory” explains how people buy assets hoping to sell them at a higher price to someone else, even if the asset itself has no inherent value. This mentality often fuels bubbles, which can burst spectacularly. I am sure you’ve read about the Dutch Tulip Mania, if not please help yourself to an amusing read here and here.
AI Bubble or Real Deal?:
The AI market holds undeniable promise, but is it currently overvalued? Let’s look at past bubbles for comparison:
Dot-com Bubble: The Internet revolution was real, but many companies were wildly overvalued. While some thrived, others crashed. – Crazy story about the dotcom bubble
Housing Bubble: Underlying factors like limited land contributed to the housing bubble, but speculation inflated prices beyond sustainability.
Cryptocurrency Bubble: While blockchain technology has potential, some cryptocurrencies like Bored Apes were likely fueled by hype rather than utility.
The AI Bubble’s Fragility:
The current AI boom shares similarities with past bubbles:
Rapid Price Increases: AI stock prices have skyrocketed, disconnected from current revenue levels.
Speculative Frenzy: The “fear of missing out” (FOMO) mentality drives new investors into the market, further inflating prices.
External Factors: Low interest rates can provide cheap capital that fuels bubbles.
Nvidia’s rich valuation is ludicrous — its market cap now exceeds that of the entire FTSE 100, yet its sales are less than four per cent of that index
The Coming Downdraft?
While AI’s long-term potential is undeniable, a correction is likely. Here’s one possible scenario:
A major non-tech company announces setbacks with its AI initiatives. This could trigger a domino effect, leading other companies to re-evaluate their AI investments.
Analyst downgrades and negative press coverage could further dampen investor confidence.
A “stampede for the exits” could ensue, causing a rapid decline in AI stock prices.
Learning from History:
The dot-com bubble burst when economic concerns spooked investors. The housing bubble collapsed when it became clear prices were unsustainable. We can’t predict the exact trigger for an AI correction, but history suggests it’s coming.
The Impact of a Burst Bubble:
The collapse of a major bubble can have far-reaching consequences. The 2008 financial crisis, triggered by the housing bubble, offers a stark reminder of the potential damage.
Beyond the Bubble:
Even if a bubble bursts, AI’s long-term potential remains. Here’s a thought-provoking comparison:
Cisco vs. Amazon: During the dot-com bubble, Cisco, a “safe” hardware company, was seen as a better investment than Amazon, a risky e-commerce startup. However, Amazon ultimately delivered far greater returns.
Conclusion:
While the AI boom is exciting, it’s crucial to be aware of potential bubble risks. Investors should consider a diversified portfolio and avoid chasing short-term gains. Also please be wary of the aftershocks. Even if the market corrects by 20% or even 30% the impact won’t be restricted to AI portfolios. There would be a funding winter of sorts, hire freezes and all the broader ecosystem impacts.
The true value of AI will likely be revealed after the hype subsides.
In the world of technology, engineers are not just cogs in a machine; they are the builders, the dreamers, and the ones who solve the problems they see in the world. And sometimes, those solutions turn into billion-dollar businesses. This is the story of the “Palantir Mafia,” a group of former Palantir employees who have left the data analytics giant to found their own startups, just like the famed “PayPal Mafia” that produced companies like SpaceX, YouTube, LinkedIn, Palantir Technologies, Affirm, Slide, Kiva, and Yelp.
1. Introducing the Amazing People from Palantir
The “Palantir Mafia,” akin to the renowned “PayPal Mafia,” comprises former Palantir engineers and executives who left to tackle meaningful problems with technological innovation, creating substantial impact and wealth. Unlike ex-consultants from firms like McKinsey, BCG, or Bain, these tech leaders leverage their deep technical expertise to solve complex issues directly, resulting in profound advancements and successful ventures.
Key Figures and Their Ventures
Alex Karp – Palantir Technologies
Former Role: Co-Founder and CEO
Company: Palantir Technologies
Focus: Data analytics
Market Penetration: Widely used across government and commercial sectors
Market Penetration: Manages over $2 trillion in assets
Capital Raised: $325 million
Raman Narayanan – SigOpt
Former Role: Data Scientist at Palantir
Company: SigOpt (acquired by Intel)
Focus: Machine learning optimization
Market Penetration: Utilized by top tech companies
Capital Raised: $8.7 million (before acquisition)
2. Engineers Make Better Founders in the Tech Industry
Unlike ex-consultants from big 3 who may excel in strategy and communication but often lack the technical depth to truly understand the intricacies of building a tech product, these ex-Palantir engineers come armed with both the vision and the technical chops to bring their ideas to life. They’ve spent years wrestling with complex data problems at Palantir, and they’re now taking those hard-won lessons to solve new challenges across a wide range of industries.
Engineers bring a problem-solving mindset that focuses on creating practical, scalable solutions. This technical acumen has allowed former Palantir employees to launch transformative companies that push the boundaries of what’s possible in various industries.
3. Market Penetration and Success of Palantir Alumni
The success of these Palantir alumni is evident through their market penetration and revenue. For instance, Palantir Technologies itself is a major player in the data analytics field, with a revenue of $1.5 billion annually. Affirm, led by Max Levchin, has made significant inroads in the consumer finance market, generating $870 million in revenue in fiscal 2021. Anduril Industries, founded by Palmer Luckey, has secured substantial contracts with the U.S. Department of Defense, contributing to its $200 million annual revenue.
Other successful ventures include Blend, with its deep partnerships with major financial institutions, and Addepar, managing over $2 trillion in assets. These companies not only showcase the technical expertise of their founders but also highlight their ability to penetrate markets and achieve substantial financial success.
4. Engineers vs. Consultants: A Compelling Argument
The technical depth and problem-solving mindset of engineers make them particularly suited for founding and leading tech startups. Their ability to directly tackle complex problems contrasts with the approach of ex-consultants from firms like McKinsey, BCG, or Bain, who often focus more on financial and operational efficiencies.
While consultants excel in operations-heavy startups, where strategic planning, financial management, and operational efficiency are paramount, engineers thrive in tech startups that require innovative solutions and deep technical expertise. The success stories of the Palantir alumni underscore this distinction, demonstrating how their engineering backgrounds have enabled them to drive significant technological advancements and build successful companies.
Conclusion
The Palantir Mafia’s engineers have leveraged their technical expertise to create innovative solutions and successful ventures, driving significant impact across various industries. Their ability to tackle complex problems directly contrasts with the approach of ex-consultants from firms like McKinsey, BCG, or Bain, who often focus more on financial and operational efficiencies. This technical depth has enabled these former Palantir employees to become influential leaders, pushing the boundaries of technology and innovation.
Disclaimer: I have immense regard for Elon Musk, so much so, I have gifted Ashlee Vance’s book on the Paypal Mafia boss to multiple people. I have even compared him to the fictional Peter Weyland in my discussions with fellow nerds. But this article is not about the trailblazing leader who sows the seeds of interplanetary exploration/colonisation. It is more about the God Complex and a reflection of multiple things that come with it which may or may not be positive. I have written about the superhero style of leadership earlier as well and briefly, touched on it in a previous article, albeit on a much smaller scale.
Introduction:
Off lately, there is a view, it is called the “superhero” theory of leadership. In which, the individual vision, charisma, and brilliance of a CEO “makes or breaks” a company. This view is absolutely dangerous — not because CEOs don’t matter or that smarts and vision don’t help. It’s dangerous because of what it ignores. Great leadership takes both mundane “management” skills and highly specialised “Domain-specific” ones. The most effective leaders have the knowledge and softer aspects that are specific to their company and industry that allow them to not just motivate, but also drive other people in the organization to do what’s necessary to succeed.
It’s been a hell of a time for the last two or three months, for news reporters, management consultants, professors in Ivy league & Red Bricks and the new age “gurus” and “influencers”. The fiasco with FTX is an unbelievable story of lapse of controls, comparable to Enron, Daewoo and Satyam. Which supposedly is an understatement as per the new executive appointed to steward it through bankruptcy. Elon Musk’s bid to takeover Twitter and the ensuing drama is equally newsworthy, not sure how much popcorn was sold to watch this drama. And finally, the end arrived for the Theranos story, with Elizabeth Holmes and Sunny Balwani sentenced to 11 years in prison.
All of these stories have something in common, they combine a very flashy leadership style with a blatant disregard for actual management practices.
The issues at FTX are too numerous to even list as a bullet point in this article, but the crux of the problem is simple. It is a complete lack of checks and balances. Plain vanilla management or accounting is not something that gets you on the cover of Fortune or The Economist or Economic Times, but oversight of a company’s activities and checking on finances is the trait of good management and leadership. At FTX, it seems to have been completely ignored. How could the company grow so much in the absence of any basic management systems? The sad part is investors and customers were also fooled by those flamboyant “leadership” (remember Nikola ?)
The trajectory of Musk’s Twitter takeover is even more disturbing. Again, it is a story of a CEO who is proud of his blatant disregard for the basics of management and an almost untainted faith in his “superpowers” and the unchallenged position of his leadership and intellect, also called The God Complex.
At the reinvented Twitter under Musk, there seems to be no regard for basic HR practices as well. Musk has massive challenges to rally and retain his employees; even assuming he wants to “rightsize” by encouraging resignations, his eccentrics may have pushed even the folks whom he intended to retain.
What can we learn from these companies? They are both ongoing, but thus far it seems that these firms have fallen victim to an all too popular belief that “superhero” leadership trumps boring management.
This is wrong, in at least two ways.
First, there is enough evidence that boring management matters and it is a source of competitive advantage for companies that take it seriously. A 2012 research by HBR has shown that management practices vary quite a lot within industries and around the world — and that companies with good management are significantly more profitable. Secondary research has confirmed that good management improves firms’ performance.
What is good management? There’s no single, comprehensive answer. But it looks like this in practice, target-setting, rewards, and monitoring. Well-managed companies set reasonable, strategic goals; set their staff up to contribute to them; and measure their progress.
Call it boring or mundane if you like — it is good business.
A major gap in the superhero theory is that it super simplifies what good leadership is. Consider the current debate over Elon Musk. To his fans, Musk’s success at Tesla, SpaceX and PayPal makes him a great leader. To his critics, the maelstrom at Twitter proves the opposite.
A major gap in the superhero theory is that it super simplifies what good leadership is. Consider the current debate over Elon Musk. To his fans, Musk’s success at Tesla, SpaceX and PayPal makes him a great leader. To his critics, the maelstrom at Twitter proves the opposite. That’s too binary, black or white. The reality is a million shades of Grey in between Black and White! Prior research does show that CEOs do matter to a company’s success, but their contribution is about more than just grand vision and raw intellect. And how much of what depends very much on the organisational backdrop.
We think of a leader’s contribution to a company along three dimensions.
Leadership Facets
The superhero narrative simplifies the entire facet of leadership on vertical differentiation, because it’s fun & easy to argue over and write cover stories about. The other two factors — Horizontal differentiation and Force Multiplier (ability to influence an organization) — are much harder to discuss and not that fun to write about.
But, when an entire generation has grown upon SuperHero movies where it is the Vision, Grit and perseverance of the Hero that saves the day, it is too hard to think in other ways. Even in pop culture, I come from a generation, where the hero gets battered and relies on the collaboration and cooperation of his friends and partners to make it out. Then, there are the “Wise old men” or the occasional woman, who gives some much-needed advice and insights.
How would this three-dimensional assessment differ from the superhero story when it comes to Elon Musk and Twitter? It would complicate the debate that both his fans and his critics seem to be having and instead would go through the three factors mentioned above. Rather than arguing solely about whether Musk is a good CEO in general, we can ask whether he has the skills and experience necessary for running a social media platform — and whether he’ll be able to motivate and manage the team that’s in place.
It’s perfectly reasonable to think, for example, that Musk is an above-average CEO, not particularly well suited to running a social media platform, whose behaviour in the run-up to his Twitter takeover ensured he would not be able to influence the people that he needed to in order to succeed.
This view of leadership is harder to put on magazine covers, and it is therefore often forgotten. But ignoring the complex relationship between leaders and their organizations is bad for investors, consumers, and ultimately for managers and CEOs, too.
