There is nothing new in the market, and the current AI frenzy reminds me of NFTs Author: Market Participant Compiled: Deep Tide TechFlow Deep Tide Introduction: As the new wave of AI proxy craze sparked by OpenClaw and Claude Code swept through social media, the author keenly sensed a crazy atmosphere similar to the NFT era of 2021. This article analyzes how social media amplifies technological narratives, how Wall Street indiscriminately sells due to the bias of "AI killing software," and why giants like Salesforce and ServiceNow are still wrongly killed by the market after delivering astonishing results. The author believes that we are in the "middle game" of a great revolution, and all the extreme optimism and panic are trying to overdraw the unfinished end in advance. The wave of enthusiasm for OpenClaw and Claude Code reminds me of the hysteria of the NFT era. The emergence of new technologies is accompanied by practicality, while also generating cultural and narrative resonance in the spirit of the times. Just like every technology that captures collective imagination at the right time, it is being processed through the same 'twisting machine' - the one that once turned JPEG images of monkeys into a $40 billion asset class. The pattern is identical: true innovation arrives, and early adopters discover true value. Subsequently, the social layer took over everything - suddenly, the conversation detached from the technology itself and turned into a performance about "standing in line". Claiming that 'this is the future' has become a symbol within the industry. Writing guidelines, thinking pieces, and exaggerating the value of the current situation can gain social recognition. The compound interest growth rate of the viewpoint even exceeds that of the technology itself. (I guarantee that there will be a viewpoint on the financial market later). Cognitive distortion machine X makes the situation worse. Social media is increasingly seen as a legitimate lens of reality, and it bends the image of facts. The highest volume sounds are not representative - they are performing 'firm beliefs' to reward the audience for such behavior. Every mainstream platform operates on engagement, and engagement rewards are extreme. 'This is very interesting and useful' will not be widely spread, 'This changes everything, your job is going to be lost' will be. A hundred retweets saying 'this changed everything' are not signals, but echoes. Echo is mistaken for consensus, consensus is mistaken for truth, and truth is mistaken for an investable theory. Girard would definitely show off her skills if she saw this scene. When enough people perform a 'faith performance' on a certain outcome, the performance itself will be confused as evidence supporting that outcome. The NFT era has proven this point: people don't want JPEG, they want 'what everyone else wants'. What is real? The latest model's capabilities are astonishing - far more impressive than NFTs, which have almost no practical abilities except for speculation and cultural signals. I use these tools every day. They have improved my efficiency in a concrete and measurable way. The underlying model is indeed impressive, and the trajectory of improvement is very steep. When I compare what I could do with these tools six months ago and today, the increment is enormous. Moreover, the broader potential is infinite. AI assisted programming, research, analysis, writing - these are not hypothetical use cases, they are happening and creating real value for those who use them. I don't want to be the person who despised the Internet in 1998. This is not the focus, I am very optimistic about AI in the long run. The key is the timeline and the gap between potential and current situation. What's not real yet - Claude won't immediately catalyze social unrest. This does not mean that humans no longer need interfaces to manage their work. This also does not mean that Anthropic has won the AI war. Think about what the most breath holding views actually require you to believe: Enterprise software - decades of accumulated workflow, integration, compliance frameworks, and institutional knowledge - will be replaced within a few quarters rather than years? Is the seat based billing model dying overnight? Will a company with annual revenue exceeding $10 billion and a gross profit margin of 80% disappear from the world just because a chatbot can write a function? [2] Wedbush's Dan Ives bluntly pointed out that "companies will not completely overturn past billions of dollars in software infrastructure investments just to migrate to companies such as Anthropic and OpenAI. Jensen Huang, who has more reason than anyone else to advocate for the disruptive power of AI, called the concept of "AI replacing software" the "most illogical thing in the world". The people who are most proactive in announcing 'Endgame' (thanks to @ WillManidis for making the word popular) are often the ones who can benefit the most from your 'unwavering belief': fan base, consultation tickets, subscription fees, conference invitations. The incentive structure rewards bold predictions that are not responsible for timing. What's interesting to me about the mirror of the market is that it makes the same mistake on the other side of the table. Anthropic released its Claude Cowork plugin on January 30th, and in less than a week, $285 billion in software, financial services, and asset management stocks evaporated. The software ETF - IGV - has fallen 22% this year, while the S&P 500 is rising. Out of 110 constituent stocks, 100 are in a loss making state. The RSI index hit 16, the lowest reading since September 2001. Hedge funds are crazily shorting software stocks and continuing to increase their positions. The narrative logic is: AI kills SaaS (Software as a Service). Every software company that charges by seat is a 'zombie'. This kind of selling is indiscriminate. Companies with completely different risk characteristics affected by AI are all treated as the same trading benchmark. When 100 out of 110 names in the index are falling, the market is no longer analyzing and is indulging in the climax of narrative. Note: Since I started writing this article, the recovery may have already begun. Pour out the bathwater and also lose the child to see what is happening inside those companies that are considered to be facing a catastrophic disaster. Salesforce's Agentforce revenue increased by 330% year-on-year, with annualized revenue exceeding $500 million and generating $12.4 billion in free cash flow. The forward price to earnings ratio is 15 times. They have just released a revenue target of $60 billion for the fiscal year 2030. This is not a company that has been disrupted by AI - this is a company that is building an AI enterprise delivery layer. ServiceNow's subscription business grew by 21%, with operating profit margin expanding to 31%, and authorized $5 billion in stock buybacks. Their AI suite Now Assist has an annual contract value (ACV) of $600 million and aims to exceed $1 billion by the end of the year. However, its stock price has fallen by 50% from its peak. Should these names be moderately downgraded in valuation due to risk? Maybe. But smart people started pricing this a few years ago. As many smarter people have pointed out, this sell-off requires you to believe both that "AI capital expenditures are collapsing" and that "AI is powerful enough to destroy the entire software industry" [11]. These two things cannot be established simultaneously. Choose one. Will some companies be truly replaced by identifying real risks? Yes. Point solutions that provide standardized single workflows are fragile. If your entire product is only built on an interface layer on top of non proprietary data, then you are in trouble. LegalZoom has dropped by 20% - for such companies, concerns are substantial [12]. When AI plugins can automatically perform contract review and confidentiality agreement (NDA) classification, it becomes difficult to defend against the value proposition of paying traditional suppliers for the same functionality. But companies with deep integration, proprietary data, and platform level foundations are completely different. Salesforce has penetrated into the technology stack of every Fortune 500 company. ServiceNow is the system of record for enterprise IT. Datadog's consumption based model means that more AI computing will directly translate into more monitoring revenue - their non AI business growth has actually accelerated to a year-on-year increase of 20% [13]. Selling digital infrastructure because 'AI kills software' is as absurd as selling construction equipment stocks because buildings are rising from the ground. We have experienced these inspiring SaaS crashes in 2022. The sector has fallen by over 50%. The median forward revenue multiple has dropped from 25 times to 7 times - lower than pre pandemic levels [14]. And during that period, the financial report performance was consistently good. The subsequent rebound was very significant - Nasdaq rose 43% in 2023. Indeed, the trigger at that time was more interest rate shocks than fundamental deterioration. The DeepSeek panic of January 2025 is even closer. Nvidia plummeted due to concerns that cheap Chinese AI models would render the entire AI infrastructure construction meaningless, but then completely regained lost ground. That fear was structurally identical to today: a single product release triggered a crisis like reassessment of the entire industry's survival. Many observers made a direct analogy between the current moment and the early stage of the Internet foam burst - technology stocks fell, while consumer necessities, utilities and healthcare stocks rose [16]. But there is one thing about the bursting of the Internet foam: Amazon fell 94%, and then became one of the most important companies in the world. The market attempted to price the 'endgame' halfway through the game, creating one of the greatest buying opportunities in history. Jim Reid of Deutsche Bank said a big truth: "Identifying long-term winners and losers at this stage is almost entirely speculative. I bet he's right. And this uncertainty - acknowledging that we don't know the outcome yet - is precisely why this indiscriminate selling became wrong. The hype merchants on the Final Fallacy X and the panic sellers on Wall Street made the same mistake at both ends of the chessboard. A group of people say that AI has already won, the future has arrived, and all institutions and job functions will be rewritten from now on. Another group of people say that AI has killed software, subscription revenue is dead, and $10 billion in free cash flow is no longer important because the business model is outdated. Both sides jumped to the 'endgame' when there were still many steps to go in the game. The gap between our current situation and technological vision will be filled by chaotic, gradual, and company specific advancements. Some software companies will integrate AI and become more powerful; A few will really be replaced; Most people will adapt - this adaptation process is slow, uneven, and not suitable for tweeted. The actual trajectory is more volatile and uncertain than implied by speculation or panic. The people who can do well from now on will be those who can tolerate this ambiguity, rather than those who are eager to grasp a premature narrative. Great managers always find a way out. Reference source [1] Girard's imitative desire theory（ https://www.iep.utm.edu/girard/ ）[2] Fortune Magazine: Why SaaS Stocks Fall Irrationally Like DeepSeek Panic [3] CNBC: The Impact of AI Tools on SaaS Software Stocks [4] CNBC: Huang Renxun Calls AI Replacing Software the "Most Unreasonable Thing" [5] Yahoo Finance: US Software Sector Evaporates $285 Billion Due to Anthropic Impact [6] Yahoo Finance: Analysis of IGV ETF Trends [7] Axios: Hedge Funds Short Sell Software Industry [8] Benzinga: Misunderstandings in Software Sector Collapse [9] Salesforce Investor Relations: Agentforce Driven Q3 Financial Report [10] Futurum Group: ServiceNow Q4 Financial Report and AI Platform Momentum [11] Fortune Magazine: AI Paradox and Irrational Analysis [12] CNBC: Software Stocks Enter a Bear Market, The sentence is:, ServiceNow and others fell sharply [13] StockAnalysis: Datadog operation statistics [14] Meritech Capital: review of the SaaS crash in 2022 [15] CNBC: Nvidia fell sharply due to DeepSeek concerns [16] Fortune magazine: Deutsche Bank talks about the analogy between the software stock foam and the Internet era [17] Deutsche Bank Jim Reid analysis report