HashiCorp co-founder Mitchell Hashimoto warns of AI psychosis
Mitchell Hashimoto, co-founder of HashiCorp, posted that companies are operating under AI psychosis, blocking rational talks on limitations and risks. Referencing the earlier shift to cloud infrastructure, he cautioned that rapid AI agent automation can produce systems registering as healthy on local metrics while growing globally incomprehensible. Investor Pete Skomoroch and founder Danielle Fong amplified the warning about favoring fast recovery over resilience.
#62Susan Zhang@SUCHENZANGapplies even for the mildly more technical neolabs, many of whom have never had the burden of building anything from scratch themselves, having thoroughly enjoyed building their own legacies on top of the pre-existing infrastructure of their previous affiliations
they don't know what they don't know, and these extremely intellectual and generally relatively smart people somehow think certain robots and their robot harnesses will solve everything beyond n=1
"but... but... anthropic! mistral!"
yes those are two exceptions because in both cases they got to either build everything or see how things were built "from scratch", with one being wildly more successful than the other having been closer to execution / been the actual execution themselves
similarly, back in the days of being a cloud infra person, there was always an aversion to hiring "experienced" FAANG engineers (more G than anyone else) since they usually all looked extremely good on paper, but could rarely reverse-engineer a half-descent cloud system since they've never had to think about primitives they take for granted
"what do you mean you can't just push a button and get more storage on the compute cluster?" - a quote from an incredulous L8+ engineer when they got to Meta and attempted to "scale" on Meta's gpu infra back in 2023
again, all very smart people with very smart resumes and pedigrees... and with surprisingly large blind spots
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#228Aidan Clark@_AIDAN_CLARK_You can believe the inner quote while realizing how wrong the outer quote is.
3-4 years ago people were amazed LLMs could answer simple questions. Now they’re complaining the critical infra they write has too many bugs. 3-4 years from now the models will be unrecognizably good.
🇺🇦 Dzmitry Bahdanau@DBahdanauin 3-4 years companies will be hiring INSANELY expensive consultants to unscrew their Mythos-created spaghetti critical infrastructure, which was 99.9% autonomous, until the 0.1% catastrophy hit don't underestimate humans. We are amazing!
@mitchellh You should have ample evidence by now if this was occurring.
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#411🇺🇦 Dzmitry Bahdanau@DBAHDANAUin 3-4 years companies will be hiring INSANELY expensive consultants to unscrew their Mythos-created spaghetti critical infrastructure, which was 99.9% autonomous, until the 0.1% catastrophy hit
don't underestimate humans. We are amazing!
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#882Danielle Fong 🔆@DANIELLEFONGwe need an alternative name to "ai psychosis" for this, for starters.
pre-comprehension bankrupcy
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
@ChowdhuryNeil cant imagine which one you could possibly be talking about
Neil Chowdhury@ChowdhuryNeilyes, and this is happening at the frontier AI labs too
@ChowdhuryNeil Agreed, in fact it's particularly strong there
Interesting to see how this may impact automated ai research
Neil Chowdhury@ChowdhuryNeilyes, and this is happening at the frontier AI labs too
#1030Herbie Bradley@HERBIEBRADLEYprobably true
living in the future has its downsides when failures are asymmetrically bad
automated AI R&D likely has *less* surface area for compounding error than SWE
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#1323Neil Chowdhury@CHOWDHURYNEILyes, and this is happening at the frontier AI labs too
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#1363Pranav Shyam@RECURSEPARADOXOn the day catastrophe hits you can’t learn your codebase instantly. Like digging a well after your house is on fire
🇺🇦 Dzmitry Bahdanau@DBahdanauin 3-4 years companies will be hiring INSANELY expensive consultants to unscrew their Mythos-created spaghetti critical infrastructure, which was 99.9% autonomous, until the 0.1% catastrophy hit don't underestimate humans. We are amazing!
#1427Pete Skomoroch@PETESKOMOROCHWell said by @mitchellh “…you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible”
Mitchell Hashimoto@mitchellhI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out. I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really). It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely. The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture. We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying. I worry.
#1702Mitchell Hashimoto@MITCHELLHI strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them. I can't name any specific people because they include personal friends I deeply respect, but I worry about how this plays out.
I lived through the great MTBF vs MTTR (mean-time-between-failure vs. mean-time-to-recovery) reckoning of infrastructure during the transition to cloud and cloud automation. All those arguments are rearing their ugly heads again but now its... the whole software development industry (maybe the whole world, really).
It's frightening, because the psychosis folks operate under an almost absolute "MTTR is all you need" mentality: "its fine to ship bugs because the agents will fix them so quickly and at a scale humans can't do!" We learned in infrastructure that MTTR is great but you can't yeet resilient systems entirely.
The main issue is I don't even know how to bring this up to people I know personally, because bringing this topic up leads to immediately dismissals like "no no, it has full test coverage" or "bug reports are going down" or something, which just don't paint the whole picture.
We already learned this lesson once in infrastructure: you can automate yourself into a very resilient catastrophe machine. Systems can appear healthy by local metrics while globally becoming incomprehensible. Bug reports can go down while latent risk explodes. Test coverage can rise while semantic understanding falls. Changes happens so fast that nobody notices the underlying architecture decaying.
I worry.