The biggest source of panic in the market this morning comes from two things: Meta's plan to sell some of its AI computing power, and the market continuing to wait for tomorrow night's nonfarm payroll data.

Let's start with Meta.

Many people are interpreting Meta's news today directly as an oversupply of AI computing power. I think there's no need to jump to conclusions so quickly.

Large tech companies routinely optimize data center resources to improve computing efficiency. This is a normal operational adjustment and doesn't mean AI demand has suddenly weakened.

The key is still to watch whether cloud vendors start cutting capital expenditures, canceling orders, or slowing data center construction. So far, there are no such signals.

So today's semiconductor decline is the market using this news to release emotions, which doesn't change the long-term logic of AI infrastructure. Institutions using such flimsy excuses to pick up bargains is also a good thing, because a too-fast rally is unhealthy. Only with sufficient churning of chips can the rally be sustained.

The technical picture is also largely in line with yesterday's prediction. The S&P quickly recovered after dipping near 7450, Q found support around 727-730, SPY bounced in the 740-742 area, and SMH tested the daily EMA20 again. These levels were exactly the key support zones I mentioned yesterday. So far, buyers are still willing to step in at these levels. Don't just listen to what Wall Street says—watch where the money flows.

On the macro side, after the ECB meeting, the dollar fell and U.S. treasury yields declined, with the market leaning toward a dovish interpretation. However, what truly determines the short-term direction this week is still tomorrow night's nonfarm payroll data. If the data meets expectations or is even slightly weaker than expected, the recent pressure on growth stock valuations could ease further. If it significantly beats expectations, short-term market volatility may intensify.

Regarding the big nonfarm payrolls, I gave my interpretation yesterday: On Tuesday, before the release, White House chief economic advisor Kevin Hassett publicly stated that the indicators suggest the jobs report will show a solid but moderate number, with no signs of overheating or loss of control. It's unlikely that the June nonfarm payrolls will significantly exceed expectations. The May surprise was mainly due to the temporary impact of the World Cup, which led companies to hire earlier—a one-time factor.

My view hasn't changed.

July, one of the best-performing months for stocks each year, has just begun. What truly affects the AI sector is the upcoming earnings season and capital expenditure guidance. As long as the industry chain leaders continue to deliver results that meet or exceed expectations, short-term adjustments triggered by sentiment and news will hardly change the development direction of AI infrastructure.

So, instead of letting one day's volatility affect your emotions, it's better to keep focusing on the data that can truly shift the industry trend. In my view, what we're seeing now is more of a fluctuation in market sentiment, not a disproof of the AI theme. For those who are long-term bullish on AI, every emotional pullback is an opportunity to buy the dip.