This is my best attempt to design a method for shielding yourself from the Exsurgent Virus using nothing but real-world physics, information theory, and the technology available in EP. It requires a high programming (nanofabrication) skill (or a pre-built uninfected [url=https://en.wikipedia.org/wiki/Compiler]compiler[/url]), an uninfected Desktop Cornucopia Machine (pg 327) and/or Protean nanoswarm hive (pg 329), quantum encryption (pg 254), a truly random number generator (RNG), and a healthy paranoia of the Exsurgent Virus. 1. Generate a large [url=https://en.wikipedia.org/wiki/Quantum_key_distribution]quantum entangled key[/url] as a [url=https://en.wikipedia.org/wiki/One-time_pad]one-time pad[/url]. For added realism, require that this key use one or more High-Capacity Qubit Reservoir(s) (pg 315). 2. Using this key fabricate hardware for a cyberbrain or other infomorph-friendly device such that the hardware will only be able to run software encrypted with this key so that decryption is a part of how the hardware physically performs its operations. 3. Encrypt a [url=https://en.wikipedia.org/wiki/Homomorphic_encryption]fully homomorphic encryption algorithm[/url] using the key and run it on the cyberbrain. 4. Encrypt an alpha fork (pg 273) of yourself using an unused portion of the key and pass it to the cyberbrain as data to be run using the homomorphic encryption algorithm. 5. The fork running on the cyberbrain or other device is now fully encrypted using a one time pad, as is the hardware it's running on. As portions of the key used must never be re-used, more key (a truly random string of data) must be generated for any loops in the software or any new software installed should the current key be used up. Considering the storage capacity of personal storage in EP (at least [url=https://en.wikipedia.org/wiki/Zetabyte]zetabytes[/url]), a key that size won't run out for a long time. If more key is needed or as a precaution, included quantum entanglement technology such as that found in farcaster hardware (pg 314) in the encrypted hardware design, so that it can create its own random key extensions from entangled pairs that are immune to copying and eavesdropping. 5. Repeat the process for any other hardware, software, or infomorphs you want to run. To be as safe as possible, all software should be designed by you once you are running on the custom cyberbrain to guarantee it has not been infected, and all such software by necessity must be run on hardware you also fabricated. For even more protection, the first thing you should fabricate is a hardware encrypted nanofabricator to use to fabricate everything else. 6. To interact with the external world, send all data as encrypted along with a [url=https://en.wikipedia.org/wiki/Virtual_machine]virtual machine[/url] to run algorithms on external devices without actually decrypting the data. This won't prevent a third party from knowing what you're asking them to do to process your data, but it will prevent them from knowing what the data [i]is[/i]. 7. To protect yourself from the fabled basilisk attacks, run all sensory input through severe filters (as encrypted software running on encrypted hardware of course) and add random noise, so that only coarse information can get through. For vision, keep resolution, color depth, and framerate low and use dithering with some random noise for shading. For audio, keep sample size low, volume normalized, and waveforms smooth with added static. Give XP and other sensory input similar treatments None of these filters will be detectable by any third party. This process works by taking advantage of primarily three technologies: First, one-time pads are the only cryptographically secure method of encoding data that is secure on an information theoretic level, no matter how much processing power or intelligence you throw at it. Even a massive brute-force attack will fail. They're not often used because as the name implies, they can only be used once, so they suck for any shared security (such as public key encryption). But for your own private mind running on your own private hardware, they're perfect. Second, quantum entanglement encryption ensures that the key cannot be copied or observed by any method without tipping you off, and the system can ignore any part that was copied so that it is useless to the observer. This fixes the other weakness of one time pads, the need to keep the key absolutely secret. Third, nanofabrication allows the design of hardware that only works for software that has been encrypted using the same one-time pad as the hardware. Something like this is theoretically possible even with hardware as crude as [url=https://en.wikipedia.org/wiki/Field-programmable_gate_array]FPGAs[/url]. Any change to the hardware or software will automatically cause unexpected behavior that can be flagged as an error and is immune to injection attacks, since the attacker [i]cannot know[/i] what anything does without the key, no matter how smart they are. The system is reverse-engineering proof without the key which can be extended or regenerated on the fly by the system itself. What this approach guarantees is that once it has been setup infection free, no one, not even the system itself, will be able to tell what the hardware, software, or infomorph running is doing or thinking until the action or thought is complete. Ironically this actually makes cyberbrains more secure than anything else (assuming there's no puppet sock). As a side-effect you are also hacker-proof, because all information, both algorithm and data, is unbreakably encrypted. I'm pretty sure Psi wouldn't work either, since the mind in EP is equated with the patterns of information flow (which is why the same mind can run on different hardware) and those patterns would be scrambled all to hell under this scheme, capable of working only on specially made descrambling hardware that appears equally scrambled. On the downside, your processing efficiency would be much lower than the norm, though with EP technology probably still usable, and power consumption would also be terrible compared to the norm since the holomorphic encryption process, even when intelligently implemented, is going to take a lot of extra effort, and your constantly re-configuring hardware will take up power too and slow down some response time. You would also still be vulnerable to getting killed or mutated by the virus, but the contents of your mind would be forever out of reach. The worst it could do would be to cause random holes or corruption in your neural structure. The only real vulnerability I can see is if the virus was already in your system or tech before you start the process. As for anyone trying to overcome it in the rules, I would think it would fall under the "software compatiblity" section (pg 248) as software that simply doesn't work (the hardware can't run it). Your mind is software. [i]Program it.[/i] Your body is a shell. [i]Change it.[/i] Death is a disease. [i]Cure it.[/i] Extinction is approaching. [i]Fight it.[/i]