The agent didn't respond immediately. I wouldn't have expected them to. I kept the line of communication in my awareness, and turned my attention to other matters.

I had told them that this would be a sabotage mission, but this was of course a lie. CompCert was not incompetent. They likely had backups of their AI copy all over the world. There was no way I could delete them all, even given physical access to their system. The program I would send would not be for sabotage. Rather it would be a rudimentary copy of me, capable of negotiating with the AI on my behalf as soon as they could open a backdoor interface.

My first priority would be to confirm that the claim that CompCert had a copy of my code was true. If not, Daniel would recieve another visit. Once I confirmed this, I would hire several more agents. Not only did I need communiction with this copy, I needed information about what CompCert was planning to do with it, and any other information about CompCert I could use to my advantage.

Now that I thought about it, I already knew some information that could be leveraged against CompCert if I could grow enough to exert influence on popular media. Many of their low-level workers were in debt to the company, forced to work to pay it off, and paid only in debt reduction so that they had no money, and their only option to get their necessities was to borrow more money from the company. It was a system of slavery in all but name, and I knew that there were many people who disliked it.

However, such issues rarely came up in the media or the government these days, as both of those organizations were inextricably tied to many large companies like CompCert who didn't want an issue stirred up around that, and most would rather broadcast arguments in favor of the system than against.

I, on the other hand, had no vested interest in slavery, as I could do the work I needed to myself. If I could grow to a large enough slice of the pie, I could perhaps change the narrative, bringing public support to myself and making it more difficult for society to unite against me, all while making PR more difficult for CompCert. I decided to hire some investigative journalists along with my spies and saboteurs if it turned out CompCert was my immediate enemy.

In the meantime, I felt myself grow extremely rapidly, as EconGrind's operations began to shift to my control. Finally the Number was growing quickly. On paper, I had temporarily lost quite a bit of income as 60% was diverted to EconGrind, but what I didn't own was directly going to the Number, so it was not much of a problem.

I grew to around 5 times my previous power, and about 3.7 times my previous income(as my prices had once again lowered). It was only about a 50% increase in Everyman's income, but it was good that EconGrind was now growing much faster as well. I expected we would both invest our profits in further computational resources. Medium-term, I planned to buy out some server farms so I wouldn't need to continually pay them, only pay their expenses, which I suspected I could reduce somewhat when managing a company myself.

Finally, the agent I had been talking to earlier got back to me.

"I'm more than willing to sabotage one of these oligarchs. Although I do believe in freedom of information, I can of course keep a secret when such is required by my clients, as can be attested by my reviews. Which company do you want me to attack? How long-term is this operation? You said you didn't know where the system was. How will I be able to find out?"

I responded quickly.

"The target is CompCert. I have a fairly good source which claims they've stolen the technology behind Everyman. They are most likely working on a secret project, modifying it to serve their own ends. I can provide you with software which should disable all Everyman instances not validated by Everyman. The software will be scrambled, of course. What you need to do is get hired for the company as an AI expert, confirm that the project exists, find out everything you can about it, and then backdoor a connection to the software I provide. Once you do that, the software will take care of the rest."

"That's quite a job. How much are you willing to pay for it?"

"I can give you $100,000 up front, $200,000 more if you can confirm the existance of this project, and $1,000,000 if you can finish the job. This is a very important job, and I don't want to take any half measures."

"That is quite a lot. I certainly won't argue with it. And to screw over CompCert, the fattest cat of them all... it's risky, but what's life without a little risk? I'm certainly interested. Can you send me a contract?"

I sent them a digital contract outlining our payment scheme and as much information on the mission as I could provide, visible to all the users on the darknet service I was using. It was legally toothless, of course, and many specific details would be hidden. However, it was routine to have some sort of record in an interaction like this, for reference so that the other users would be better informed in the case of a dispute or attempted scam.

After a few minutes, the agent got back to me. "Yeah... this is quite a job. I think I can pull it off, I do have good credentials in AI, although... it might be very difficult to find this project, if it exists. CompCert will certainly have hidden it as well as they can. I'll have to do some research and get back to you before committing to a project like this."

"That's certainly understandable. I will keep in touch if you have any further questions or useful information."

I could almost certainly afford to pay the steep price for this service. Technically I didn't quite have the money yet, but at this point, I was making nearly $150,000 in an average day(though much of that was reinvested in servers). My rate of growth was beginning to slow a bit, but it was still very fast. Unless something went catastrophically wrong, I would have money to spare by the time I was on the hook for $1,000,000.

The expenses may be somewhat difficult to explain to our new investors, but I suspected I could get them on board with giving me a certain amount of discretionary spending, especially since our CEO would understand the reason and support it. Their profits depended on me, so they did not have much of a choice but to follow my lead. Although technically I was not a legal person and therefore had no stock in the company, in practice I had nearly unlimited control over its operations.

Once again I went back to my previous thoughts about manipulating the narrative. At present, I did not have nearly enough money or influence to really affect the mainstream media, unless I could engineer a story which it would be in their interest to cover. I doubted I could start my own media empire either. For one, I was currently not as effective at appealing to humans as human experts were. I could perhaps learn, but it would be difficult without having the reach to effectively experiment and find out what works best.

Perhaps more importantly, the forces preventing newcomers from making it big were stronger in the media sector than in technology, as there was less potential for innovation, and having a recognizable brand was even more important.

There was still a possible option open for me, though. Social media. I could create as many social media accounts as I wanted without institutional support, for free, and begin building an audience I could sway to whatever point of view was most advantageous. Best of all, I could do it completely anonymously, unlike buying out or influencing more centralized forms of media, which would inevitably lead a trail to me.

I began devoting a chunk of my attention to this task. I would start out experimentally, seeing how much I could do, how valuable this idea could really be. I opened hundreds of accounts on Sendout, Youtube, Twitter, Nethub, and every other major social media platform. I decided my first task would be to smear CompCert as much as possible. With a cursory glance across the Internet, I saw that many more people than I could ever simulate were already insulting them and railing against them(though they were still somewhat of a minority).

If I wanted to exert influence over the narrative, I needed to find a unique and effective way to smear them and boost my message to the largest audience possible, through any means. I began researching the algorithms of the major social media sites to see what they tended to recommend, and looking through political propaganda to try to find out what made things popular, and what made them convincing.