저자: PattiPoppe · 외국인
요약 보기
PG&E CEO 패티 포프는 회사의 펀더멘털은 탄탄하지만 캘리포니아 산불 책임 관련 입법이 교착 상태에 빠져 있어 20억 달러 규모의 자본 지출을 삭감하고 일부 프로젝트를 연기할 수밖에 없다고 말했다. 그녀는 간단한 입법 조치만 이루어지면 잔여 리스크가 제거되고 유동성이 확보되어 투자등급(investment-grade) 신용등급 취득이 가능해지며, 이에 따라 수익과 배당이 늘어날 것이라고 주장했다. 또한 특별회기 소집을 촉구했고, 법안이 공개되면 통과 가능성이 높다는 신호가 될 것이라고 밝혔다. 이와 별도로, 이 코너에서는 AI 에이전트가 컴퓨팅 수요를 증가시키면서 메모리, 전력, 데이터센터, 엔비디아(NVDA) 등 인프라 분야가 수혜를 입고 있다는 점을 다뤘다. 아울러 규제 리스크에도 불구하고 AI 투자 사이클은 여전히 유효하다는 주장도 소개되었다.
다듬은 글
and capital reduction in dollars. We have a capital plan of $11.4 billion after that $ 2 billion cut. So, it's not zero. We're not going down to zero, but we've cut it by $2 billion, and that will delay the start of housing construction. This will delay the connection of renewable energy to the grid.
This will delay our, uh, connection of new and growing loads here in California. California is growing, we are powering it , and we need access to cheap loans and quality investors to make it happen. And look, we're not done yet. Speaker Rivas postponed the bill yesterday, as you suggested. He said they wanted to get back to work.
They have already scheduled several hearings. Governor Newsom was adamant that this work must be done. Uh, we really hope they can find a way to get back into negotiations and, uh, finish the job for our clients. Californians are waiting. Well, did they know about how you put all these shares in a trust and you did all this...You had to hurt the shareholders .
There was nothing you could do. You explained very clearly what would happen. So it wasn't like someone came and said, "I'm shocked." Because you laid it all out, but I thought you had already done the hard work. We did the hard work. You're not wrong, Jim Kramer. We did the hard work. Look, the fundamentals of this company have never been better.
Our base of operational safety, wildfire safety, reliability, availability—this structure exists. My people did the job for California. Now we need our legislature to finish the job so we can take it to the next level and let this horse out of the stable. We have a great company serving Californians, and we just need a simple legislative solution to make it happen .
Now reporters are downgrading the stock. I think they just feel like I can't, I just can't deal with it. Because, you know, I wanted to say just buy them, but I spent a lot of time on it myself and came back thinking, "I really hope Patti can pull this off." I truly believe the governor is for it, but this California legislature, from the day I followed it as a reporter in 1978, has been very difficult.
Well, look, they've done a lot of hard work over the years. They have actually created an industry-leading wildfire liability system, but there is a real residual risk that has spooked investors, and it can be fixed. This residual risk can be eliminated and the liquidity issue resolved with simple wording, a simple solution.
The current legislature needs to build on what has been done in the past. The governor led this process for 8 years. I know he clearly expressed his position. The California Earthquake Administration report was clear. Our Public Utilities Commission was clear. Everyone agrees that the status quo does not suit consumers .
Let's do this, and then investors will benefit from unlocking this value to customers, lowering rates, reducing the cost of our debt, and ultimately achieving investment grade . Okay, tell our viewers what it means to be investment grade. What would that allow you to do? Well, that would allow us to put that $2 billion back into the plan.
This would allow us to increase profits by 9% or more every year . This would allow us to continue to increase our dividend. This would allow us to become one of the leading utility companies in the world. I serve the best and smartest clients. I serve Northern California, the Bay Area, Silicon Valley, the Central Valley, the great breadbasket of the world.
I have the Diablo Canyon nuclear power plant . I know this is one of your favorite places. I was there on the day it was put into operation. These assets are waiting, they are trapped. We are like a racehorse in the stable, waiting to be released to serve Californians at a lower cost, with greater reliability and safety.
But we need to resolve this legal framework once and for all. Well, you're here because you're a trusted person in the utilities industry since the days of CMS. Um, data centers. Do you still...I know you had a place for them. I know if done right, it ca n't be bad. It's just that many are not as smart as you and didn't know how to benefit other ratepayers.
How are you doing with this? Everything is going great. Our plan for 2030 includes 1.8 gigawatts. This has been maintained despite the reduction in capital expenditure. Although capital expenditure cuts will slow this beneficial load growth. It's a simple pricing issue, Jim, regarding data centers. If you set the right price, it can be beneficial to everyone else's rates.
We set prices correctly in California. We have capacity in the network, ready to use. Now we need to provide a system of accountability that will allow us to raise capital to do our great work, bring in new customers, and lower rates for everyone. Just make sure that this is exactly what we are talking about in the context of wildfire legislation.
This could literally be happening right now. Maybe something will happen tomorrow, right? Just because the legislature is closed doesn't mean it's all over. They can hold a special session. They can hold a special session. I hope they do it. This will take several days, as the bill must be published 72 hours before the vote.
But this will be a good signal for investors: if the draft law is published and meets the minimum requirements, the probability of its adoption is very high. But first they need to publish the document, call a special session, and make sure we complete this work. We are so close to this. We are so close to this.
And I just think that under the leadership of Gavin Newsom and Speaker Rivas, they will really be able to get things right for our utility customers. I've been following your career. I've been following your career the whole time . I've been following your career since you took over this company and saved it.
They saved her and saved the customers a lot of money. I have to believe they will let you win. They should let you win. This is Patty Poppe, CEO of PG&E. Patty, thank you very much. Artificial intelligence research may be entering another important phase, and this time the focus is shifting from simply building larger models to having AI agents actually perform tasks.
The key question investors are asking is: will demand for AI computing power accelerate again? One argument is that enterprises are moving from experimenting with AI to actually implementing it in business, and this transition can . The claim that an AI agent can require thousands of times more computation than a simple AI interaction with a single response is discussed.
If this happens on a large scale, the amount of infrastructure required could be enormous. That's why investors are once again scrutinizing the AI infrastructure ecosystem. Companies involved in memory, power chips, and data center infrastructure are tied to this cost cycle. Names like Sandisk, Lumen Energy, and Nvidia were highlighted as part of a broader infrastructure story.
The main thesis is simple. More AI agents, more actions, more inferences, and more computations, and therefore potentially much greater demand on infrastructure. Nvidia CEO Jensen Huang has also strongly opposed the so- called AI apocalypse narrative. His position is that the idea of humanity being destroyed by artificial intelligence by 2030 is unrealistic.
He argued that scaring people with extreme scenarios is unnecessary and irresponsible. Huang also says that Nvidia has a direct economic interest in showing that AI products are implemented safely and productively. If the technology stopped delivering secure value, the entire industry would ultimately suffer.
But the debate over AI safety has created another major uncertainty— regulation. Before the recent successes of AI companies, many investors assumed that the US would continue to take a fairly soft approach. This assumption has now become less certain. Some leading AI companies are asking the government to become more involved in the process, while also proposing forms of self-regulation.
The question is how much additional oversight will eventually emerge and whether it will change the pace of AI development. Some investors still believe that the long-term supply- demand imbalance in AI infrastructure remains extremely large. Customer demand continues to grow, and businesses are increasingly using AI.
And model vendors continue to capture huge amounts of computing power. So, even if regulation becomes stricter, the argument is that AI development is unlikely to simply stop. Model launches, infrastructure investments, and stable demand for computing are expected to continue. One interesting example comes from the field of AI-based investing itself.
One investment firm says its model reduced the impact of AI trading earlier this year. At one point, the firm's exposure to AI-related assets was around 30%. Over time, the model reduced this figure to about 15%. In mid- August, the impact rose again to around 30%, and the explanation was that the models are designed to distinguish between short-term market noise and long-term fundamentals.
And the model obviously didn't react chaotically to every headline around AI regulation. This is an interesting difference between human investors and algorithmic systems. People can see scary headlines and react immediately. AI models can be designed with a long-term investment bias in mind. The firm's argument is that fundamentals remain unchanged and demand for AI is still explosive.
Computing needs continued to grow, and large customers continued to spend heavily. NVIDIA management also pointed to very strong future chip shipments. Discussion of a potential doubling of chip shipments, along with expectations of significant revenue growth, reinforce the idea that the infrastructure cycle remains strong and there could still be a significant amount of catch-up spending ahead as enterprises increase their use of AI.
But there is another development in this story—a new AI product from Meta, Moose. What makes Moose interesting is not just the model itself. This is an opportunity to bring generative AI closer to ordinary users.
Comments 0