In recent years, large models have emerged as a significant branch of artificial intelligence, drawing substantial venture capital interest and consistently capturing the attention of investors in the capital marketsHowever, as the fervor surrounding the so-called "battle of the models" begins to cool, an increasing number of investors are raising concerns over the potential financial bubble that large models may pose, signaling a need for the market to take heed.

A report published by the European Central Bank has identified the risk of “bubbles” associated with stocks related to artificial intelligenceIf investors harbor excessively optimistic expectations for the AI industry, a significant downturn may occur once those expectations fail to materializeSimilar sentiments have been echoed by domestic experts, who argue that, much like numerous past technological waves, as large models move beyond their initial phases, a bubble seems almost inevitable.

Reflecting on the past year, the cost of training these large models has been on a steep upward trajectory

Data from research institutions indicates that the costs have escalated from a few thousand dollars initially to hundreds of millions today, with projections suggesting that the next generation of large models could incur training expenses reaching into the billionsFaced with such exorbitant costs, even robust large corporations find it challenging to sustain these endeavorsFurthermore, the specialized teams required to train and maintain these models demand exceptionally high levels of technical expertise, often accompanied by substantial salary expenditures, leading many nascent startups to recoil in the face of such overwhelming barriers.

From a commercialization standpoint, the monetization of large models remains an unresolved issueInvestors within the venture capital landscape have expressed concerns about the reasonable valuation of large models, pointing to significant premiums and uncertain prospects for commercialization

• Capitalizing on Underperforming Assets in the Recovery
• Global Oil Giants Announce Sudden Cut!
• What Happens if Bank Deposit Rates Drop to Zero?
• Will the Fed Extend Rate Cuts?
• Boosting Consumption to Drive Growth

The reality confronting AI companies globally is stark: massive investments in technological development yield meager commercial returnsThe capital market, driven by profit motivation, may quickly withdraw from funding related to large models if tangible returns are not promptly realized, rendering a bubble burst merely a matter of time.

Some investors once viewed artificial intelligence as the leading edge of future innovation, believing that merely being positioned within this domain and successfully training large models would guarantee abundance in technological dividendsHowever, the diverse range of application scenarios has resulted in many high-profile models encountering criticisms regarding their reliability which led to their performance falling short of expectationsThis, in turn, has dampened the excitement and enthusiasm among subsequent investors.

In the short term, professionals working within the artificial intelligence sector face the critical challenge of swiftly translating the research outcomes of large models into tangible economic value

This requires identifying clear paths to monetization for investors, avoiding the trap of mere theoretical proposals that lack substanceIt demands an equal focus on the practical applications of technology alongside the pursuit of advanced innovations to ensure that developments can effectively transform into real productivity, ultimately providing investors with concrete returns.

Looking ahead, large models will inevitably follow a cycle similar to many emerging phenomena, likely entering a phase of bubble burst followed by a reassessment of valueThe collapse of a bubble, often feared, can signify a turning point that allows for renewed clarity and opportunityHistory reveals that many groundbreaking new technologies emerge from the ashes of a bubble collapse, having undergone persistent iterative upgrades, ultimately attaining a form of value that eclipses previous standards.

Moreover, individual investors should recognize that while large models represent a substantial aspect of artificial intelligence, they are not the sole avenue for development