One-sentence takeaway: AI is not a fake story, but the market is starting to demand that it “show its homework” — revenue, cash flow, and real users. Nothing less will do.
Yesterday, WikiBit Weekly Space AMA, hosted by Vicky, brought together Cube CEO Merlin, SlideFun‘s Evelyn, Openverse’s Florian, and Vocasa co-founder Abdul to discuss one of the hottest questions in the market right now: Is AI trading driving real growth, or are we witnessing another market bubble？
AI investment is pouring aggressively into chips, cloud computing, data centers, and new AI models. At the same time, inflation is running hotter than expected, interest-rate concerns are returning, and valuations are coming under pressure.
So investors are starting to ask: Is the AI high-speed train being powered by real earnings and adoption, or has market excitement simply raced ahead of the fundamentals？
The discussion also has direct implications for Crypto. The AI narrative is influencing token markets, AI agents are beginning to participate in trading and payments, and capital is moving among technology stocks, Bitcoin, and AI-related digital assets.
1. When CPI Runs Hot, Does the AI Train Start to Shake？ The Real Problem Is “Proof of Value”
U.S. September CPI came in hotter than expected, with headline inflation rising 3.7% year over year and 0.6% month over month, with energy being one of the contributing factors. The market immediately began asking: Could interest rates stay higher for longer？
Florian believes AI stocks are particularly sensitive to interest rates because their valuations depend heavily on future earnings. The best-case scenario would be moderate PPI growth combined with healthy but not overheated retail sales. The worst-case scenario would be hot PPI + strong retail sales, which the market could interpret as a signal that rates may need to stay higher for longer.
What about Bitcoin？
Bitcoin could initially be sold like other risk assets. But later, it could potentially benefit from concerns over monetary credibility. AI tokens, meanwhile, could be even more volatile because they generally have weaker fundamental support and lower liquidity.
Merlins view is even more direct: An AI company can be doing well while its stock still falls because the broader economy is working against it.
Under high inflation, smaller companies face higher borrowing costs and greater financial pressure. Larger players such as OpenAI, Anthropic, and Meta have much deeper financial resources and can withstand the pressure more easily.
But what investors should really focus on is: How is the budget being spent？ What does free cash flow look like？ Are revenue expectations actually growing faster than expenses？
Merlin went even further, arguing that if a company raises $1 billion and spends all of it on infrastructure without generating revenue, it starts to look like “money laundering” from his perspective.
Evelyn added that inflation is a trigger, but it is not the entire story. The markets real concern is whether AI valuations have become too high and whether massive infrastructure spending can generate sufficient returns.
That means PPI should be watched for signs of broader inflationary pressure, while retail sales should be monitored for signs that consumer demand is becoming too strong.
AI stocks could come under pressure, while Bitcoin and AI tokens may be more heavily influenced by liquidity conditions and interest-rate sentiment.
In one sentence: CPI may be the trigger, but AIs real challenge is “proving its value.”
2. AIs Biggest Enemy May Not Be Inflation — It May Be the Speed at Which It Burns Cash
If there was one keyword repeatedly emphasized throughout the AMA, it was cash flow.
Large technology companies can rely on massive financing capacity, strong capital-market support, and substantial cash reserves. That means they can continue investing heavily in AI infrastructure and tolerate losses for an extended period.
But what happens if an AI companys revenue growth consistently fails to keep pace with its infrastructure spending？
Merlin offered a very straightforward assessment: Investors should not look only at the revenue growth reported by AI companies. They also need to examine how much free cash flow remains after infrastructure spending.
Because:
Revenue growth does not automatically equal a successful business model.
Imagine a company whose revenue is growing rapidly, but almost all of that revenue is immediately reinvested into infrastructure. In reality, the company remains highly dependent on investors continuing to provide capital.
The company may be growing, but its business model may not have truly proven itself.
Going forward, AI companies need to prove four things simultaneously:
People are genuinely using AI.
Users are willing to pay for AI.
AI can consistently generate revenue.
Revenue growth can eventually outpace cost growth.
If a company can prove only the first two, it has demonstrated product-market fit.
If it can prove the first three, its business model is beginning to work.
But only when it proves the fourth does AI truly move from being a “technology story” to a “business machine.”
3. Big Tech Earnings Season: Dont Just Flex Your AI Muscles — Show the Free Cash Flow
The upcoming Big Tech earnings season will put Microsoft, Alphabet, Amazon, and Meta under the spotlight.
The question is no longer simply:
“How much are you investing in AI？”
The real question is:
“How much money is AI actually making？”
Florian offered a useful checklist:
Microsoft: Is AI strengthening Azure？ Are customers actually paying for Copilot？
Alphabet: Are cloud growth and AI monetization working, without damaging its core search business？
Amazon: Is AI demand for AWS remaining strong while margins hold up？
Meta: Is AI improving advertising performance and user engagement enough to justify its massive infrastructure spending？
And the most important metric:
Free cash flow after infrastructure spending.
Revenue growth matters, but investors increasingly want to know:
How much more money do you have to burn to achieve that growth？
Meta is scheduled to report earnings on October 21, while Microsoft is expected to report on October 28.
For AI-related Crypto projects, these earnings reports are likely to influence market sentiment more than fundamentals — unless a project has a genuine commercial relationship with one of these major companies.
Evelyn put it vividly: Simply spending more money on AI is no longer enough. The market wants to see AI become a real business.
In the short term, strong earnings could bring attention and capital back into the AI narrative.
In the long term, however, AI tokens still need to prove their users, demand, and utility.
Merlin once again emphasized free cash flow: if revenue grows but all the additional money is immediately consumed by spending, the business model remains essentially static.
Big Tech should reduce spending expectations, focus on revenue growth, and prove that users are genuinely willing to pay for AI tools.
Abdul added a business perspective: AI is a revolutionary technology, comparable in transformative potential to electricity and the Industrial Revolution. It will reshape many industries.
But companies cannot adopt AI simply because everyone else is doing it.
They need to solve real problems and generate measurable returns.
In trading, for example, AI can serve as an execution tool: humans define the strategy, while AI handles execution. The advantage is that AI does not experience emotions and therefore does not engage in revenge trading.
4. A New Model Does Not Automatically Mean New Revenue: From Headline to Impact
The AI industry is developing at extraordinary speed.
New models, new agents, and new products are appearing almost constantly.
But simply launching something new is becoming less impressive.
What is truly scarce is this:
People who keep using it — and are willing to pay for it.
When evaluating an AI project, investors should not focus solely on how advanced its latest model appears. Instead, they should ask:
Does it have real users？
Are users actually staying and using the product？
Are enterprise customers genuinely reducing costs？
Can revenue cover expenses？
Is the project overly dependent on external financing？
Is its infrastructure spending reasonable？
Funding rounds, partnerships, new models, and technological breakthroughs can all create short-term price rallies.
But these headlines do not automatically translate into long-term value.
A project announces that it has raised a huge funding round. The market gets excited.
But investors should ask:
“The money has been raised. Now what？”
A project launches a new model.
The real question should be:
“How many people are actually using it？”
A project announces a partnership with a major technology company.
The real question should be:
“How much actual revenue has that partnership generated？”
That is why evaluating AI projects requires a shift:
From Headline to Impact.
And that may be the key to distinguishing market hype from genuine value.
If you want, I can also turn this into a more native English crypto-media style version for WikiBit, with stronger headlines, shorter paragraphs, and wording optimized for Google/AI search citation.
5. AI + Crypto: Payments and Trading May Be the Strongest Use Cases for Agents
AI agents are increasingly seen as one of the most important use cases for Crypto. But which directions actually have practical value？
Merlin is bullish on payments, monitoring trading positions, tracking large numbers of tokens, and coordinating liquidity across different platforms. But he also pointed out that the trust problem has not been fully solved.
When evaluating an AI + Crypto project, investors cannot look at the narrative alone. They need to examine the actual trading activity of real user wallets, whether the product genuinely needs a token, whether customers are retained, and whether businesses are actually saving money rather than simply increasing spending.
Evelyn believes payments and trading are the strongest use cases. If an AI agent can help users save money, save time, and complete on-chain payments, then it has genuine value.
The real signal of adoption is not “How many wallets are there？” but:
Will users come back after the incentives disappear？
Will they continue trading？ Will they keep using the product repeatedly？
Florian added that AI agents could pay for data and services, handle small-value commercial transactions, monitor wallets, identify risks, prepare treasury operations, automatically generate reports, and simplify cross-chain interactions.
Users should not have to understand every blockchain or every bridge.
The key requirements are:
Reliable data + clear permissions + transparent records.
To determine whether adoption is real, look for repeat users, recurring transactions, continued willingness to pay even without token incentives, and continued growth after rewards are reduced.
Abduls criteria are equally practical: look at active users, usage frequency, retention, and willingness to pay.
If a project provides decentralized computing power, AI agent services, or machine-learning models, investors should ask whether real customers are actually paying and receiving measurable value.
They should also examine sustainable revenue, whether the token has genuine utility, and whether demand remains when the token price falls.
One point deserves particular attention:
The strategy should be designed by humans, while AI handles execution.
One of the biggest problems Crypto traders face is not necessarily a lack of analytical ability. It is emotion.
AI does not experience fear or greed, and it does not engage in revenge trading after suffering a loss.
So the real value of AI in trading may not be “predicting the next candlestick.”
It may be:
“Helping traders strictly execute the trading discipline they have already established.”
Another major opportunity is reducing the complexity of Crypto for ordinary users.
Wallets, gas fees, different blockchains, cross-chain transfers, bridges, DEXs, token approvals — none of these are particularly user-friendly.
AI agents could take on a new role:
the intelligent operating layer of the Crypto world.
Users would not need to understand which blockchain to use or which protocol to interact with. They could simply tell the agent:
“Complete this transaction for me.”
The agent could handle asset identification, route selection, cross-chain transfers, execution, and confirmation in the background.
But there is one major prerequisite:
Trust.
If an AI agent has access to a wallet, funds, or trading permissions, a single mistake could have consequences far more serious than an ordinary AI chatbot error.
For AI agents to truly enter the financial sector, three core issues must therefore be solved:
Reliable data + clearly defined permissions + complete traceability.
6. How Can You Tell Whether an AI + Crypto Project Is Really Hot — or Just “Hot Wallet Numbers”？
Crypto projects can very easily create the illusion of growth.
For example, a project may claim to have 1 million wallet addresses.
That sounds impressive.
But there is an obvious question:
Are those wallets actually users？
Or are they a combination of airdrop farmers, bots, project-owned wallets, incentive-program addresses, and one-time users？
That is why the more important metric is not simply how many wallets exist, but whether users come back.
Real user growth should look like this:
Level 1: People use the product.There are genuine transactions and real interactions.
Level 2: Users come back.They do not simply interact once and disappear; they continue returning.
Level 3: Users are willing to pay.Even without token rewards, airdrops, or subsidies, users are still willing to pay for the product.
Level 4: Revenue is sustainable.The project generates genuine and recurring business revenue.
Level 5: The token has real utility.The token is not simply a fundraising or speculation tool; it plays an important role in how the product actually operates.
So, genuine growth is not about the number of wallet addresses continuously increasing.
It is about users becoming increasingly dependent on the product.
7. If AI Cools Down, Where Will the Money Go？
If inflation data or Fed comments push yields higher and reduce the attractiveness of AI stocks, where could the capital go？
Florian believes the answer depends on why AI is falling.
If inflation is high and yields rise:Energy, gold, and defensive stocks could benefit.
If economic growth weakens:Cash, government bonds, and defensive equities may become more attractive.
Bitcoin is more complicated.
During a broad risk-off event, Bitcoin could initially fall because investors need liquidity. But later, if the market begins to believe that the macro environment favors scarce digital assets, Bitcoin could recover.
There is an important difference between rotation and a broad sell-off.
Rotation means capital moves from AI into another relatively strong sector.
A broad sell-off means stocks, Crypto, and commodities all decline together while the dollar and volatility rise.
In the short term, AI tokens are likely to follow the broader Crypto market more closely because liquidity conditions and Bitcoin's direction remain powerful drivers. The AI narrative may simply amplify the volatility.
Merlin does not fully agree with the simplistic idea that “money will flow from AI into Crypto.”
His argument is straightforward:
The AI token market is relatively small and is still part of the broader Crypto industry.
Without a risk-on environment, even positive AI news may not be enough to push AI tokens sharply higher.
If investors simply believe AI valuations have become too expensive, capital could move into energy or traditional businesses instead — because both Crypto and AI assets remain highly speculative.
Crypto may first need to see a stronger global economic environment and a return of risk appetite.
8. AI Cooling Down Does Not Mean AI Crypto Automatically Gets the Money
This is particularly important for Crypto investors.
There is a common assumption:
AI stocks fall → money leaves AI → money flows into Crypto → AI tokens rise.
But the guests believe capital does not move that mechanically.
The reason is simple:
AI Crypto is still Crypto.
If the broader Crypto market is not receiving sufficient capital inflows, AI tokens will not automatically attract money simply because AI stocks are falling.
In fact, the opposite could happen.
AI tokens generally have smaller market caps and weaker liquidity. As risk appetite declines, they may become even more sensitive to market-wide selling pressure.
Therefore, in the short term:
AI tokens are more likely to follow Bitcoin and the broader Crypto market than simply track AI technology stocks.
9. A “Health Check” for AI Projects: A Practical Investor Checklist
Based on the views shared throughout the AMA, the entire discussion can be condensed into a simple investment framework:
1) Does AI actually solve a problem？
Do not start by asking, “How advanced is this AI technology？”
Start by asking:
“What problem does it actually solve？”
2) Are the users real？
Do not focus only on registrations, wallet counts, or social-media followers.
Pay closer attention to active users, retention, and repeat usage.
3) Are users willing to pay？
This is one of the most important tests.
If the subsidies disappear, will anyone still pay to use it？
4) Can revenue outgrow expenses？
Pay particular attention to the relationship between revenue growth and infrastructure costs.
If revenue grows 10% while costs grow 50%, that does not necessarily represent a healthy business model.
5) Does the AI product actually need a token？
If the product can function perfectly well without a token, then ask:
What value does the token actually provide？
6) Are users still there after incentives end？
Airdrops, points, and rewards can attract users.
But the real test of a project's sustainability is:
Do those users come back after the rewards disappear？
7) Does the AI agent provide genuine execution value？
A truly promising AI agent should do more than chat.
It should be able to:
Understand → Decide → Execute → Verify
This is particularly important in areas such as payments, trading, cross-chain operations, and asset management.
10. Conclusion: The Real AI “Wealth Code” May Not Be the Next 100x Token
The AMA ultimately offered a very clear answer:
AI is neither simply a “bubble” nor simply a “bull market.”
It is entering a more important phase:
From telling stories to showing results.
In the past, the market was willing to pay a premium for models, funding rounds, valuations, partnerships, and narratives.
Going forward, the market may increasingly focus on:
Users. Revenue. Profit. Cash flow. Real-world adoption.
For AI + Crypto, the standards will be even higher.
A genuinely valuable AI project needs to answer two questions simultaneously:
What problem does theAI solve？
And:
Why does it actually need Crypto？
If both questions can be answered clearly, AI + Crypto could have enormous long-term potential.
But if the only answer is:
“Because AI is hot, so we have AI too,”
then the project may simply be another player riding the next narrative cycle.
For ordinary investors, the most important question may therefore not be:
“What is the next AI token？”
Instead, remember this:
Dont be fooled by the AI story. Look at what AI actually creates.
When the market shifts from “telling stories” to “doing the math,” projects with real revenue, real users, real cash flow, and real utility are far more likely to survive.
The AI train is still moving. But passengers are starting to have their tickets checked.
The projects most likely to survive the cycle will not necessarily be the ones that shout “AI” the loudest.
They will be the ones capable of turning AI into revenue, cash flow, and genuine adoption.
And that may be the real dividing line between the next phase of AI and Crypto.
Add a stronger opening hookCondense repeated market-rotation points