Transcription
In 2025, 64% of all venture capital in the US has been allocated to artificial intelligence startups. Venture capital investment in artificial intelligence hit $50 billion in the second quarter, accounting for nearly half of all VC funding. Companies like Open AI were valued at $300 billion without even being profitable or publicly traded. OpenAI at 300 billion. Is is that too expensive? Is it is it a deal?
In 2024 alone, Google, Amazon, and Meta spent over $400 billion on AI infrastructure or startups. A $20 billion dollar pledge from Amazon to build artificial intelligence innovation campuses. However, 70% of these funded AI startups still do not generate real revenue. This has raised alarms among some experts who claim that the AI boom could be turning into a bubble like the dot bubble of the early 2000s. Hype cycles come and go. The metaverse, web 3, crypto. What about the mother of all tech cycles, the dotcom bubble and burst? So, right now, are we in the middle of another one with the generative AI and chat GBT?
So, how AI became the new .com bubble? The .com bubble was an economic phenomenon that took place between the mid 1990s and the early 2000s. During that period, the mass adoption of the internet sparked a surge of investment in web-based companies, just as is happening now with the AI boom. The NASDAQ composite rose nearly 400% from 1995 to March 2000, only to crash 78% by October 2002, essentially wiping out all the gains from that period. Venture capitalists and speculators invested without considering real business models, and many companies went bankrupt after burning through their funding without generating profits.
Today, the world is experiencing a new technological boom, the explosion of AI. Over the past decade, we've become surrounded by AI systems that perceive our worlds, that support our decisions, and that mimic our ability to create. Since the public launch of Chat GPT in late 2022, investor attention has surged. Global funding for AI startups rose from $18 billion in 2014 to $119 billion in 2021. And by 2023, generative AI accounted for nearly 30% of that total investment in the United States. In the first half of 2025, investments in AI represented 64.1% of the total value of venture capital transactions in the US. This has triggered warnings and comparisons to the dot bubble.
The enthusiasm for artificial intelligence has sparked an unprecedented wave of investment. For example, OpenAI was valued at $300 billion in April 2025 without even being publicly traded. Anthropic, its direct competitor, is seeking to raise an additional $5 billion despite its products not yet generating revenue. Amazon is reportedly considering deepening its ties with the AI startup Anthropic through a new multi-billion dollar investment. In the first quarter of 2025, the four largest venture capital deals in the US were AI related, totaling $26.6 billion. This capital concentration is reminiscent of the dotcom bubble frenzy when funds flowed to companies with no finished products solely because of the allure of emerging technology.
Big tech companies have responded to this euphoria with an arms race to acquire talent and infrastructure. AI engineer salaries have skyrocketed with compensation packages reaching $100 million, especially in Silicon Valley. So, OpenAI CEO Sam Alman, he says that Meta is trying to poach his top engineers offering what he called $100 million sign-on bonuses. Google, Microsoft, and Meta are projecting a combined spending of $45 billion on AI infrastructure by 2026. This figure doubles previous estimates and represents a jump of over 13% compared to the previous year. Alphabet, for example, announced it will allocate more than $85 billion annually to its artificial intelligence projects.
This has had a ripple effect on financial markets. Currently, the tech sector accounts for 34% of the S&P 500 index, an even higher proportion than during the peak of the dotcom bubble in 2000. But despite the enthusiasm, most generative AI products have yet to demonstrate a truly large-scale revolutionary impact. While they may be useful tools, their transformative potential is still up for debate. We're not saying here that AI isn't leading to be productive. We're just saying that AI tools, yes, they're here and they're being used. The payoff may be more uneven than the hype suggests.
Many users report that these models provide inaccurate, shallow, or outright incorrect responses. In practice, generative AI has proven more effective in supporting repetitive tasks than in disrupting entire processes. This has led to a disconnect between the futuristic promises made by companies, such as replacing 80% of administrative jobs and tangible results. Moreover, much of the value of these startups is based on expectations of future revenue that have yet to materialize. Revenue projections are in many cases speculative. According to data from CB Insights, more than 70% of AI startups that received funding in 2023 and 2024 still do not generate operating profits. Added to this is the fact that talent and early stage company acquisitions by large corporations have increased by 40% in 2025. In other words, returns on investment remain uncertain, raising doubts about the sustainability of this speculative wave.
In addition, the productivity gains from AI are being questioned. Tech giants like Microsoft and Google are outsourcing more and more coding to AI in a productivity push, but some new research shows the tools might not be as helpful as some expect. Another troubling symptom is the disconnect between stock prices and economic fundamentals. Nvidia, one of the largest AI chip suppliers, lost nearly 17% of its market value in a single day after the release of an open-source model promising similar results. Although Nvidia has greatly benefited from the AI boom, this event showed how volatile current valuations are where future expectations can collapse with an unexpected competitive breakthrough.
On top of that, there is a risk of financial contagion. The excitement over AI has pushed other speculative assets upward, including cryptocurrencies, echoing the asset correlation phenomenon during the .com bubble. Back then, the NASDAQ crash dragged down nearly 1,500 small tech companies and had a systemic impact on global markets. Down 1.7% here, a loss of 37 points or so. Apple shares are just getting hammered this morning. We're down by between 3 and 4 1.5% generally across these markets. Let's talk about the speed with which we are watching this market deteriorate.
Today, a negative event in the AI ecosystem could trigger a similar domino effect given that many institutional portfolios and venture capital funds are heavily exposed to the sector. On the other hand, regulatory and social ecosystems could also limit AI's future growth. Various entities including the European Union and US authorities are working on regulatory frameworks that could impose restrictions on data use, privacy or copyright. A clear example is the lawsuit filed by Disney and Universal against the generative image AI platform Midjourney. This is the first time major Hollywood studios have sued an AI company for copyright infringement, marking a legal turning point in the industry. Disney and NBC Universal, which is CNBC's parent company, are filing a joint suit against AI company Midjourney, alleging copyright infringement.
Skepticism is another factor, as about 55% of Americans believe AI will be just one more technology among many without a decisive impact. Restrictive regulation combined with a decline in public and consumer enthusiasm could trigger a sharp shift in market perception. At the same time, the amount of money being invested with projections estimating up to $7 trillion by 2030 raises questions about the sustainable return on that investment. As with the dotcom era, it is possible that many AI companies will not survive the current speculative cycle. Some will be acquired, others will disappear, and only a few will emerge as sustainable leaders. [Music]