We’ve all heard of enshittification:
“ Sloppification” isn’t as well known, but we’ve all also experienced (and are living through) some of our once beloved products becoming riddled with AI slop. Sometimes this is in the software itself. Other times, it’s the content around it.
Both processes deliver similar effects. Certain products and services used to perform at, and be experienced at a certain level of quality. In both cases our experience of these products and services has degraded.
Blippification
I think there is a new 3rd thing that we’re in the early innings of. Over the last few months, and continuing for at least the next year or 2, many services we’ve come to depend on will face an onslaught of probing by hackers/agents. This is the other side of the coin of software having become much easier to build; it’s also much easier and cheaper to probe for vulnerabilities, and exploit them. The probing, hacking, and responses (ie the companies patching critical systems) result in what I call “blips”
You already see this in the last few weeks as various systems have experienced blips. A blip is a temporary, often unconventional degradation of a service we’d otherwise not even think of as a target. A blip can range just from a website not working as expected, to a previously dependable system becoming completely unusable for a period of time.
A hallmark of a blip is that it’s often unexplained, poorly explained or presented with conflicting explanations. This isn’t malice; in a lot of cases the explanation reflects the evolving understanding of the entity experiencing it.
Reality has surprisingly many failure modes
Blips feel most acute when they extend into the physical world. Think of flight stoppages, fridges overheating and other infrastructure degrading that you don’t think of as software. At this point almost all infrastructure relies on control systems that are software based, mostly written by humans (with the attendant vulnerabilities).
I posit that the sources of more blips than we want to admit are driven by either explicit attacks on these services, or probing enabled by agents, or patching activity by organizations reaching at human speed to vulnerabilities being exploited at machine speed.
Blips degrade the experience for everyone in 2 ways
Explicit: a lot of services will either go down, or decide that certain features are not worth the risk, and just deprecate them. In both cases, anything built on those features/anyone using those features will feel like the product got worse. The benefit they experience from those products will be lower.
Implicit: a lot of services will be either a) down or b ) slow, and you’ll just not know if it’s degraded because they screwed up, or because they are actively being probed/attacked and are responding a that moment in time
Even thoughtful makers with good security practices will experience blips as agent enabled attackers throw more and more tokens at cracking them open, more continuously, with newer, more powerful agents, and ever newer vulnerabilities.
The blippening
But in both cases the result is the same . . .we’ll just experience more and more blips across more and more services. We will experience failure modes in systems we didn’t know could fail.
One second order effect is how uneven this will be. Some entities will be able to withstand the blips that come their way, harden themselves continuously, and survive. Some will not. For some the cost of handling this new environment will outstrip the margins of the business long term. In the worst case scenario, some entities will go out of business because they simply cannot react fast enough to a) the disclosure of and b) the exploitation of vulnerabilities. [1] It’s to early to tell which is which.
Thanks to Daniele Perito for reading this in draft.
[1] Anecdotally, the consequences of getting breached is extremely lopsided in favor of large companies. Basically every public company that gets hacked experiences a brief dip in stock price and effectively no discernible impact on their business (eg see CrowdStrike)



