Unconscious AI, Salesforce Dreamforce and Quantum Wafer Runs

Hey everyone, happy Thursday. Man, it's great to be back in the studio. I was out in Denver this week at Splunk Conf, and it was a fantastic conference as Splunk and Cisco continue to come closer together, bringing observability to security and security to observability, and AI and agentics to all of it.

And I've written a bunch of, and Mike Vizard was there with me. Mike's written a few articles as well. You should check it out.

But we're here to talk about other stuff today, and we've got a great gang of members to talk about it with. Let me introduce you to our gang. First of all, we have one of our writers here, Jamie Hampton, reporting in.

Hi. Hi, Jamie. How are you?

I'm doing great today. Thank you. Good to hear it.

Joining me is my friend, Wiki Wang. Hey. I communicate more on LinkedIn with Wiki than on here, but Wiki, it's good to see you.

I hope all is good. Yeah, nice to see you in person. Yes, well, on the screen.

Yeah. And then last but not least, it's our man out in the Valley, John Schwartz. John, things are crazy, huh, out there?

Ugh. Jesus. Yes.

Every night there's a little surprise. Yes. Every morning there's a surprise.

It's a surprise in the morning. I try to- Those are the ones, yeah ... nuts.

Anyway, happy Thursday. Let's jump into it, everyone. Continuing a theme that I guess has been going on since last weekend around the whole, I'm calling it the AI scare.

AI's going to make us extinct. We've got to slow down. We've got to stop this.

We've got to lock it in. No, we shouldn't. Well, maybe we can do it slower.

The next shoe to drop here, and it's like there's enough shoes here for a spider. But the next shoe to drop is that OpenAI is now disclosing six more incidents that were previously undisclosed. John, you've got an article up around, there's a few articles we have in here, but what have you got with this?

Well, let me first of all say, can I make a sarcastic, snarky comment to begin? Sure. Because does it seem to all of us that everyone is trying to take moral high ground in AI ethics and security all of a sudden?

And I always wonder what type of influence is taking place over these leaders. Is it the investors? Is it IPO pressures?

It just seems as if everyone is trying to either assume blame and then say, "We're addressing it," or they're criticizing or pointing fingers at others to point out how flawed their systems are. And the latest example is we have the head of AI at Microsoft, Mustafa Suleyman. So he's calling out Anthropic.

That's what this story is about. So he's calling them out for training Claude in ways that acknowledge uncertainty around AI consciousness and moral status. He's arguing that encouraging AI systems to behave as though they might be conscious could make advanced autonomous systems harder to control or shut down.

So he's pointing this at Anthropic. It's interesting, during his screed, he also complimented Anthropic's CEO for being one of the more responsible CEOs in this area. So he's talking out of both ends of his mouth.

At the same time, what is not helping matters, or what is accelerating the you're to blame, I'm to blame chorus is these incidents. And you mentioned, Alan, OpenAI last night disclosed six new incidents of unexpected or concerning behavior by its AI models. And it brought up the word misalignment, which is now trending very well among the popular press and among our politicians.

There have been so many instances or incidents that have been reported, it begs this question of what are the motivations all of a sudden of this happening? I guess part of the motivations are these companies have to come clean because some of their ex-employees are pointing out the issues within the companies that were never made public until now. So again, I'm a little bit skeptical, and I talked earlier this week on Monday on Techstrong gang, Tom Hollingsworth brought this up.

It's like this timing is a little curious, but it is a legitimate issue and I'm glad they're talking about it, but I also think, as you said, it's the sky is falling mentality we have right now. Well, let me ask Jamie, Wiki, any thoughts on this before I wade into it? To be honest, I think, first of all, I do agree we need to give OpenAI and Anthropic a little bit of credit on this, because earlier I was teaching AI security class at local university.

We had a red teaming lab there, and my students tried to use their lab results from the Anthropic and OpenAI. They couldn't get an answer. So they did put something there to help slow down some harmful attack or something.

But I think for this evolving area, it is hard, at this stage, to cover all of those because it's so evolving. There are a lot of emerging things we never see in the past. But I feel like for this specific one, I want to clarify one thing, this alignment is never equal to security.

Mm-hmm. There are a lot of things currently. I talked to one of my alignment friends who did research for alignment for over 10 years yesterday.

There are still alignment problem, need people to do more research and try to figure out, which is very hard. But on the other side, there are a lot of way from security perspective we can govern it, certain piece to make things work. So, in this article, what Suresh, Suleyman mentioned- Yeah.

Suleyman is developer from Microsoft. Yeah. A gentleman from Microsoft.

So mentioned, I feel like he's little bit mixed two things together. But those two, I personally feel like we need to separate the thing. Currently, what we can control definitely is at a security level.

Yep. Yeah, that's my point. Okay.

So I haven't had a chance to weigh in on a lot of this this week because I was out at the Splunk conference. There's a lot of my friends who are saying, not being snarky, but they're being realistic and saying this seems very coordinated. Mm-hmm.

Amodei asked to slow down. Elon and Sam join in, and then you got that one side, and then on the other side you've got Jensen and Prime Minister Yi of China, as well as Trump himself saying, "No, no, don't slow down. " Or President Yi, not prime minister from China.

And so the lines have been drawn. No, we don't need to stop. Yes, we should go slow.

And I think what is emerging is maybe a little middle consensus where we're not going to slow down. It's unrealistic because the whole world's not going to slow down. But what we can do is maybe can we build processes?

Can we build a public policy and oversight so that we have a methodology, so that we have a roadmap, we have guardrails that allows us to do this? And part of that disclosure by OpenAI was part of some internal disclosure policies that they are now following, which is looking for these kinds of things and making sure they're transparent about it, as well as what they're doing to make sure it doesn't happen again. Yeah.

So that's progress. But at the end of the day, I think it's also a tip to reality, which is you cannot make the world slow down We don't have that. I don't care who you are, whether you're China or the US, there's going to be rogue players.

It only takes one. And this talk of when the hugging face bot got loose from OpenAI back over the summer, there's talk that they seeded some self-replicating code. Now that they're out there anyway, you're not going to be able to.

There's all kinds of stories. We're venturing into the unknown, but we always... I wrote articles this week on it.

There were scientists who thought that when we started the first chain reaction for a fission bomb, and then again for a new hydrogen bomb, that we would have a runaway chain reaction, which would consume the Earth, right? You would have a mini sun, if you will, for a short period anyway. So there's always these kinds of things with new technologies.

And I'm not downplaying the potential of AI to do bad things, but I think we also have to be realistic about what we will and will not do, and what human nature is about, and what the drive to be first. Well, there's a-- Oh, go ahead. Yeah.

I want to support you on this because, recently I talked to a few AI founders in my podcast, and surprisingly, they're thinking about if AI goes so fast, what humans can do. Even they are the person who's on the edge, try to develop the AI. So, yeah.

I think currently it's not like we force AI to slow down. It's like if we keep AI in this speed, does something around can keep up the speed to support? Think about in the company, we're talking about AI related ROI, right?

Mm-hmm. In the infrastructure data center, do we have enough things to support? Do we have enough compute?

Even in the application part, do we have enough use case to support the whole system? That's an overall thing we should consider. But- Mm-hmm ...

from my perspective, I'm also thinking like, oh, maybe we should slow down to see if the regulation can come up more, like guidance can come up on the compliance security perspective as well. Well, that actually might be a positive outcome here because we do have this definite push. I'm not sure, I'm not confident it's going to happen, but at least there are a lot of bills being introduced in Congress to address things like data centers, kill switches, security.

And there is this dynamic between Silicon Valley, Wall Street, and the White House on trying to address this in some manner, because this really has taken on these political overtones, especially between the US and China. So I do think there is pressure on that front because midterm elections are defined in great deal by this discussion, and it is permeating throughout political coverage. It is on the front page and on- It's on the front page- It's everywhere ...

the middle page, and the back page. Right. It was six weeks ago.

Yeah. But John, to your point, I don't know how many different bills have now been proposed in Congress, none of which are going anywhere. None of which...

And it's a quagmire, right? So kill switch, for example. I think John Kennedy, the senator from Louisiana- Yeah ...

is pushing for this, and yet on this, we have Rand Paul, who's against it for certain reasons. We have a definite... It's weird, though.

It doesn't go on party lines, by the way. No. So there are disagreements across the board on how this will be handled.

The lobbying efforts from big tech are definitely there. Trump doesn't want to have anything to do with legislation, and in fact, they'll probably bury this so they can pass crypto bills and other type of bizarre bills. Well, no, I think they're on vacation for the rest of the year right now.

I don't know if they come back before the elections. They're not going to do it. Yeah, that's not going to happen.

If anything happens, I'm very doubtful it will. It would only happen around December, and that's not going to happen. No, that's a lame duck Congress.

That's a lame duck. Yeah. And it's a shame.

Look, the tech industry has always recited the mantra of, "We'll take care of our own. " But this really is a case where I think public policy has to weigh in. Mm-hmm.

I think people need to be heard. Whether you believe AI can cause human extinction or not, or the likelihood of it, is one set of facts. But I know from my security experience, an ounce of prevention is worth a pound of cure.

And putting in guardrails and processes now is going to be a lot easier than having to unravel things later. Mm-hmm. And I'll leave us with that and jump into our next topic, if it's okay, if no one else has anything else on this one.

But I don't think we're done discussing this topic anyway. It's going to be continuing to rear its head as every time AI does something new, and it's doing something new every day, as we say. So I was at SwanCon this week out in Denver, but several of our Futurum colleagues and a lot of people, I think over 40,000 was the number I saw, were out at the Moscone Center in San Francisco for this year's Dreamforce.

And as every other conference around us, Dreamforce was all about AI and agentics as well. Wiki, I don't know if you were at Dreamforce, but I know you covered and saw it on the various outlets. What was your take on this one?

Yeah. I wasn't in Dreamforce, but I think it's a good event to watch out. " That was the very interesting part.

And I do read this text drawing reporting this week regarding these events. I feel like this isn't just a pile of routine product launch news, right? I feel like we talk about Salesforce and ServiceNow last year regarding how they want to transition to the AI world, and this time, I think they made a really big signal on this.

Salesforce has an answer on the existing questions. Will AI make vertical SaaS products like CRM irrelevant? The answer is no.

We're going to change our strategy on the products and pipeline, and then make us as a big hat in the AI era. You can see the data moat they put there from the call, COA. And the governance checkpoint, the harness part, they mentioned they build up the harness.

And also the aggressive M&A to fill the capability gap with Listen Labs. You can see the sad news of the Listen Lab also turned down the next round of funding, trying to make it happen. It's a big bet on the real execution risk.

And, yeah. But I feel like they try to build up their own AI governance layer and try to make the harness portion and approach to their existing customer to make it happen, which is what it is for me, yeah. John, were you at- Yeah.

I listened in to the keynote by Mark Benioff. And you know Mark, he's been around forever. He was at Oracle, he was at Apple, he founded Salesforce.

He's the I think he is one of the best marketers in Silicon Valley, which I say as a compliment and also not as a compliment because, in a sense, he was kind of overselling. And in a sense, it was a little bit of defensive posture on his part. There was a lot of allusions to, "There's no such thing as a SaaSpocalypse.

There's no such thing. " And yet they came out with a series of measures or announcements that show they're concerned about what's going on. They introduced AI Force, which decouples CRM data for the AI era.

They debuted with NVIDIA something called Koa, which is a reasoning model targeting CRM workloads. As Ricky mentioned, they launched the Enterprise AI Harness for multi-vendor agents. Allegedly, they're in talks to buy Listen Labs for $2 billion.

And one thing that did happen that was not planned was they had an outage yesterday morning at the second- I heard it was some rogue agents from OpenAI. I'm only kidding. Yeah.

It was resolved very quickly, but it did happen early on East Coast time, but it was nonetheless a bit of a black eye, and some of their competitors brought it up. But nonetheless, this is their show. This is his pitch.

This is Benioff's dog and pony show. And as expected, Jensen is there. I think Sam Altman participated.

They had an incredible turnout. And in fact, Mitch Ashley was out here, as well as James McGuire covering this for us. Nice.

So it's interesting to me because I talk a lot to ServiceNow, and these two companies, Salesforce and ServiceNow, are always at loggerheads, and they like to snipe about each other quite a bit. So, this is the turn for Salesforce. ServiceNow has their show in May, and it's kind of a jostling between the two for the enterprise market.

It'll be interesting, but I do sense with Benioff, he's a little bit on edge. He's not going to admit to it, but he was referencing a lot of stories and a lot of reports that he disputed, market research reports that he disputed, that point out that Salesforce isn't ideally positioned and is a little bit susceptible, as well as they're doing. Things happen so fast, we just don't know.

But again, interesting timing, interesting show, but a lot of bloviating, honestly. Yeah. Agreed.

You know what? Look, I was never as close to Salesforce as you have, and Benioff as you have been, John. You've been covering them almost, I guess, since they started Salesforce.

I've known Mark a long time. I do like him personally. I knew him when he was at Oracle.

I did a story at a trade magazine where it was one of the first stories that he'd been interviewed in. So he would invite me to his apartments when they were forming Salesforce. He wanted reporters to see, and he was very hands-on, and he had all these great ideas, all these great concepts.

No one could ever argue with his success and his vision. Absolutely incredible. But sometimes you have to decouple the ideas and the hot air at times with the reality of the situation.

I don't disagree. But here, interesting thing, if you just look at their announcements with AI, and you know what? Credit to him.

He saw AI coming and embraced it. Right? They had Agentforce earlier on, right, where they were going to have a garage of agents, their own, tens of their own, thousands of their own, and they'll manage other people's agents as well.

And they came out, I forgot the name of it, was it Einstein was Salesforce- Einstein ... wasn't it? Yeah.

So they had Agentforce, they had Einstein, they had- Einstein, in a sense, was a precursor to-- That was their AI play before- Yeah ... AI became ubiquitous, yeah. Yeah.

Then they had sort of their own version of Bedrock, like their own harness, if you will, that would allow... And kind of Einstein was like that, right? It allowed you to plug into different LLMs on the back end and still go through Salesforce, and that would keep your data safe and everything else.

You don't hear a lot about that anymore. Then they did the deal with Claude just a week or two before Dreamforce, right? With Claude Force or whatever they're calling it, right?

Claude agents now in Salesforce. Now they've done this NVIDIA deal with more forces, if you will. You know the old story, if something works, you go with it.

If it doesn't, you just keep trying different things. " You're a very astute man, Alan. I think that's exactly what's going on.

They are throwing as much, I would say stuff, I won't use the other word- Yeah ... against the wall as they can, hoping it's going to stick. And some of it is, some of it isn't.

They did the Slack deal. There's the Tableau. There's the Mulesoft.

They're acquiring these companies. They're grafting them on. And in a sense, I keep bringing it up.

It really does. There is this kind of posture that they are trying to prove that they are as relevant as others and that they are on the leading edge. But they're trying so hard, because it just sends to me mixed signals.

Yeah. Agreed, man. Agreed.

I have different opinion, right? So Salesforce is one of the respectful company, and I agree with you, Mark is very nice person and respectful one. But from what I can see is currently a lot of big hats, they're trying to do what startup companies doing.

Mm-hmm. Which means that the AI product may become at a mature level. There are some strong needs there.

If those sales company, if they don't do those, they may out of the business in the future. So they try to protect themselves. They try to get into this business.

From what I can see, all their published information currently address the pinpoint enterprise currently facing. They're doing the right direction. The only thing I was thinking is like, "Oh, if they're doing the harness, if people are going to accept it," because they're saying it's a cross-platform, but it's sitting in their platform.

I don't know if people will like that. Steph? Yeah.

The company is interesting because twice they've tried to have co-CEOs with Benioff. But for Benioff, it's always going to be his show. There was a guy named Keith Block, who was co-CEO not a very long time, Bret Taylor.

And so in a sense, there were murmurings, I won't say from whom, but people I know who work at Salesforce, that Mark was disengaging, kind of his eyes kind of went off the ball because he wanted to be a celebrity CEO, and he wanted somebody else to run the company. When they didn't do it to his satisfaction, he came back, and now in a sense, he's trying to reassert that. So there is this kind of trying to prove something, and he did it before.

He did it with Salesforce after he left Oracle, so maybe he does it again. I just think it's not happening as easily for him as he thought it would, and this is more of a grind than he anticipated. Agreed.

And just something on the Tell Labs, for those who aren't familiar, Tell Labs, AI company, but using a lot of voice control and voice. And this is something else that I've been seeing. Jamie, I don't know if you've seen it, but Sam Altman mentioned it.

Mm-hmm. AI and working with agents is moving past the prompt, right? You no longer necessarily have to type everything into a prompt.

Almost "Hello, computer" from "Star Trek" from Scotty. That's sort of how we're going to interact with our agents. We may talk to them, but we certainly may not have to type with them.

I don't know, Jamie, have you seen anything like this, or? Yeah. There are plenty of people that are using generative AI through voice prompts.

I personally know people who speak to ChatGPT in their cars. They'll be driving down the road. Yeah.

And I've been in the car with someone where they have a question to ask. And you can choose from a plethora of voices and- You know with Astra, they actually have new personalities. Yes.

So I definitely think that one day we will move beyond the typing heavy prompts For sure. I think that makes sense. To give Sam Altman his due, he's not talking just replacing typing with voice.

He's saying that these agents will understand what you want better, anticipate what you want better, and so you won't have to spoon-feed them, in essence. Right? " I think we saw- I'm not going to prompt it all out for you.

I don't know, Jamie. Wasn't there a report, I got it, it's so hard to keep track of all this stuff, but that OpenAI was looking into some sort of device that's voice-activated that would incorporate Jony Ive's concept. Well, yes.

And I don't think we've heard anything about that in a while. Well, because it's been buried under the runaway AI that is going to kill us all. But I think that's next year, and that's like an Apple type of device.

Yeah. It'll probably come out around when Apple announces their next batch of new devices. And OpenAI pushed back their IPO to next year.

Yes. So maybe around that time. And so maybe there's something to the voices that say a lot of this is being very, very orchestrated.

I've been pushing my book a lot lately. I've got a great video out on LinkedIn and social media on it today. But in my book, I look at AI compared to some historical build-outs.

Two of which, the electricity build-out and the telephone system build-out. So in electricity was a man named Samuel Insull, and the president of AT&T is not Fell. I forget his name right now.

But both of them recognized that government regulation was a double-edged sword. It would cap their ability to make untold profits, but it would lock in their monopoly status. It would lock in, in essence, a golden handcuffs that would allow them to make money.

Not as much money as they maybe potentially could without it, but it would lock them in as a monopoly, as indispensable, as the book says. And Tell, I think, was the guy from AT&T's name. But anyway, this is what we might be seeing here, right?

Maybe Dario and Sam want to be locked in as the models of choice, and if the government owns a piece of them, right, as they've discussed, and they promise to be good boys and not have runaway agents, and in exchange all the competition's eliminated, including the open-weight models from China. There's a lot of conspiracy theories running around with that as well. So it's a tangled web that's being woven here, folks, for sure.

But let's move over. We're moving along. Let's move to block three.

Jamie, this is a story on quantum, which is kind of OpenAI's little brother at this point. Yeah. But one day it's going to have its day, too, and it's coming sooner than we think.

You have a story up on some new quantum silicon and wafers and chips and so forth, functionality. Yeah. IBM, they have a spin-out company, Anduron, and they received a $1 billion award from the Commerce Department.

It was finalized. This was announced back in May, and so it was finalized, and Anduron is now running quantum wafers through its 300-millimeter foundry in Albany, New York. So what makes this interesting is that we're moving quantum hardware manufacturing towards repeatable production, instead of just these one-off lab fabrications.

So it's very exciting to see where quantum hardware manufacturing is going to take quantum research, because right now they're working on superconducting qubits, but the factory is going to be adding more technologies as time goes on. So that's exciting. Absolutely.

We're seeing more and more. I just saw, I forgot what company, is selling Brazil a quantum computer for research. Got a picture of it, it was pretty cool.

And it's definitely IonQ made an acquisition last year. I wrote a special report on this a couple of weeks ago, where quantum is the first computer industry that arrived before the computer did, right? Mm.

Because the fact is we still don't have working fault-tolerant qubit quantum computing. And quite frankly, my friend John Willis was up at the IEE. The IEEE had a quantum conference, I think it was in Toronto this week.

And John was there. He's going to write a story on it for us, but I spoke to him about it. The fact is when you ask people, what is the killer app for quantum computing?

What are we going to do with quantum computing that we can't do? Right? A lot of people will talk post-quantum, quantum's ability to, let's say, break encryption, but that's not the killer app.

That's the bad guy's app, right? And we're planning against it. What is the killer app?

" Well- Yeah. Oh, go ahead, Jamie. That's the question about quantum advantage.

That's where that research comes in. It's like, what problems can we solve with more of an advantage using quantum? And right now it's physics and chemistry and these really high-level things.

So how are we going to bring that down a little as far as- Yeah, but it's basic research ... for our everyday lives. Yeah.

Right? But, again, I'll go to "Star Trek," right? What was it Scotty had to get when they were trying to put the whales in the spaceship?

Transparent titanium or something, right? It's materials research, right? And maybe quantum allows us to invent new materials and ways of doing things.

But it's kind of out there, and it's very hard to translate into something I can wrap my hands around right now. John, you were going to say something. Oh, I was going to ask Jamie, maybe, I think you probably already mentioned this, but I want to ask you a little bit more about, do you think this news represents a real-world manufacturing milestone for quantum computing?

I think that manufacturing is kind of, it's not a very glamorous problem. But it's one of the more immediate problems. Even though with the first quantum wafers actually being manufactured, this is a step beyond just announcing a foundry.

But it doesn't prove that fault-tolerant quantum computing has arrived. It's still- But it's a step. It's a significant step, right?

But it's building that manufacturing foundation that we're going to need if we're ever going to scale quantum computing at all, and if we're ever going to have enough research capability for all these different qubit models. Yeah. No, we're in, I hate to use the word, but we're in that foundational time right now where we're still basic building blocks that are needed to build this computer, right?

Again, as I wrote in the report, the industry came before the computer. The quantum industry was about a $550 billion industry last year, half a trillion, better than half a trillion dollars, and we don't have a working computer. That's a lot of money to spend, right?

And what are we getting out of it? I'm not 100% sure, other than we've done great with post-quantum algorithms. Right?

To help us lock stuff down when quantum becomes real. Yeah. So for this one, actually, I have three thinking.

" Maybe a little bit early, but worth to think about it. And second one is quite interesting if you see what IBM plays their role here, right? They're doing something for the infrastructure.

Are they going to be the quantum Anthropic in the future? What do you think? If you go back to AI, people would tell you that Watson was really the first AI, right?

Right. And then somehow no one thought Watson was going to be the new OpenAI or Anthropic. They thought those would be the new Watson.

I don't know if IBM will capitalize on all this foundational research they've done to become the industry leader. There are pure play quantums. There's one moving right down the block from us here at our Boca office, D-Wave.

D-Wave Quantum is moving from California, their headquarters, right here to Boca Raton, actually in the old what they call BRIC, the Boca Raton Innovation Center, which is the old IBM campus where IBM was building and pioneered their PCs. And it's the old IBM campus, and now D-Wave Quantum is moving in there. So interesting connection- Yeah ...

between IBM and them. But I don't know. I mean- I think a more apt comparison would be, is IBM going to become the new TSMC in the States?

TSMC, right. Because if quantum develops the way that the classical semiconductors did, we need a- Foundries ... a shared fabrication kind of infrastructure, rather than every company building its own foundry.

" That would be interesting. Good for IBM. Yeah, because I remember earlier this year in the RSA, right, IBM is the only company bring the quantum staff into the RSA.

Mm-hmm. " Yeah, NIST and MITRE are working, but again, that's just post-quantum algorithms. That's not truly quantum computing per se.

It's algorithms that hopefully quantum computing won't break, so we can still encrypt data. But different, and while that has generated business, that's very different than building quantum computing and qubits and so forth. Anyway, hey, we're about out of time.

This was a great show today. Happy Thursday. Jamie, Wicki, thank you for joining us.

I know, Jamie, you've been coming on a lot recently and writing, and thank you very much. John, it's always good to have you on with me, man. Thanks, man.

I appreciate it as always. Keep doing great stuff. I hope you out there have enjoyed our discussion today.

We'll be back here tomorrow with our Friday gang show. As usual, we do this Monday to Friday, noon every day, Eastern Time. tv, our YouTube Techstrong TV YouTube channel, and our Techstrong TV OTT channel, which is on Android and iOS.

If you're running the new iOS 27, it runs on all the iOS. But as well as Amazon, Roku, and Apple TV. On top of that, it does run on our websites as well.

I wanted to give a shout-out, though. We did this last year with our friends at Digicert. We put together a list of what I think we call the Quantum Security 25, top 25 people in quantum and quantum security, and we're doing it again this year in partnership with Digicert.

So at the end here, we're going to have a little announcement of that. So don't go anywhere. But until tomorrow, this is Alan Shumel from Techstrong.

Thanks for watching. We have to be ready for an unexpected break. But the really important stuff that's existential threats to our businesses has to be done now.

And firmware, code signing, all those types of things need to be protected as soon as possible. The problem is, it's already the last minute, right?