Principal AI PM @ Amplitude
Joined October 2025
- Tweets147
- Following16
- Followers631
- Likes366
Pinned Tweet
(1/6) i'm the pm on Wave at @Amplitude, our agent for self-improving products. for the past few months we've had it running on our own product. it finds problems, ships fixes, and measures whether they worked. here are some highlights:
just spent 10 minutes ranting a master plan to @WisprFlow just to realize i didn’t fully press my fn key and nothing was captured
humanity will eventually look back on its treatment of non-human animals as one of the greatest moral failings of all time
incredible tech demo. there’s a lot of haters saying “realistic human speech is so inefficient, why would anyone want this?”
if you can’t imagine a use case for ai voice beyond maximizing raw information throughput, you simply don’t understand what most of humanity’s relationship to ai will look like over the next decade
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
robots have good hands now and tl;dr is that the pinky finger did not make the cut
this is great news - we should continue to find ways to subtly handicap humanoid robots to maximize our chance of survival
the people saying “humans will find a way” are making $1M+ TC and are deeply out of touch with the average human experience
if even 10% of the population becomes permanently disenfranchised to this level, we will face civil unrest like we’ve only seen once before as a country
there’s something perverse happening in the recruiting landscape right now
for whatever reason, candidates shoving every conceivable AI buzzword into their resume is…working
i think this tactic has found a terrible evolutionary niche:
- resume-filtering platforms like Ashby do simple keyword searches instead of more substantive, holistic scoring
- this makes a naive sourcer feel like they have the perfect candidate (100% match on keywords in job description)
- the candidate then gets filtered out in a phone screen, or almost certainly rejected after first HM call
it’s just successful enough that under-qualified candidates get interviews, but ultimately leaves them little chance of being hired
waste of both the candidate’s time and the hiring team’s time
i genuinely think teams (including my own) should introduce a pre-phone screen step which is passing the resume through chatGPT for “buzzword/bullshit” detection, amongst other tests
i’ve never run this prior to a call (to not bias myself) but i always do afterwards and the models are 95% aligned to my own intuition and conclusions
if you are a PM on the job hunt:
unfortunately, you have been trained for decades to think about your job as an "operations” role. you aligned stakeholders, hosted rituals, ran enablement, unblocked engineers, triaged customer feedback, etc
that's process. if you find yourself explaining your processes during an interview, things aren't going well
my favorite question to ask candidates to cut through the noise is:
"why is the product/feature you worked on more successful because you worked on it versus any other PM?"
you immediately get past any tailwind effects (working at early pmf company where all metrics looked amazing), teamwork effects (marketing had a killer launch sequence), or generic "process" answers
candidates have to talk about what they believe their unique contributions were to a product's success. with the right person, it's the start to an incredible conversation
plug: if you think you got an exceptional answer to that question, we're hiring ambitious, AI-centric senior & principal-level PMs here at @Amplitude_HQ to work directly on our AI capabilities and strategy
we are bringing the technology for self-improving products to the world, and i'd love to work with more folks who are passionate about that vision
jobs are posted online!
babe wake up, anthropic just found another plausible way we might all die
anthropic.com/research/self-…
critical milestone: >50% of the product improvements that Wave comes up with are scored positively (worth shipping)
only 18% are thrown away
for a moment, compare this to the best human PMs you’ve worked with
Wave can drive absolutely insane impact at scale, even at this acceptance rate
but progress isn’t stopping - stay tuned
(1/6) i'm the pm on Wave at @Amplitude, our agent for self-improving products. for the past few months we've had it running on our own product. it finds problems, ships fixes, and measures whether they worked. here are some highlights:
turns out everyone likes identity-based hiring
finally my most impressive accomplishment can be recognized… being born here 🦅
Vice President JD Vance announces the suspension of the PERM Program for Microsoft:
"Thanks to an ongoing investigation, we have decided to suspend the PERM program for Microsoft, which means they will no longer be able to take those H-1Bs and apply for permanent resident status within the United States of America. Now our message to Microsoft and to any other foreign or domestic corporation is quite simple. We obviously want you to thrive in the United States of America. We want people to invest in the United States of America, particularly the American technology sector. We’re the biggest boosters of your industry. We want you to continue to grow and to employ a lot of people, but we want you to employ American workers. And so our message to Microsoft is, you’re great American company, but you’ve got to hire great American workers. You cannot lay off American workers and then replace them with foreign indentured servants."
waymo this, waymo that. in an 18 minute lyft i learned that my driver christopher:
- was married twice, but still best friends with his first wife because of their son
- juggled multiple girlfriends at the same time after his divorce
- met his second wife at a club, couldn’t get pregnant together, and started ivf
- struggled to produce a sperm sample at the ivf clinic because he heard the nurses in the other room giggling
- has stage 4 liver cancer but is trying to stay alive to see his daughter get married
- believes i should have kids asap because the government is trying to replace native-born citizens with immigrants
all of that without me asking a single question
ai will never replace this experience
reminder: a human didn’t trigger this investigation or point the agent at a known issue
this was an agent first understanding how users engage with the product, then figuring out what drives retention, then designing and shipping an elegant solution that improved the user experience and makes them come back
like holy fuck we’re really doing it folks
Today we're releasing the first instance of self improving products with a customer: @PixAI_Official
Wave found that publishing artwork on PixAI's AI-image generation product was correlated with retention, but most new users never discovered the publish button. Wave suggested prompting users to publish immediately after generating an image. Now, when a user runs out of image generation credits, a tooltip points them to the Missions tab, where they can publish.
Results:
- Mission opens increased 76%
- Publish rate increased 10%
- Day 1 retention increased 3%
We're opening Wave to more teams who want to spend time with it. Let me know if you want to try it for your product!
“Thou shalt not make a machine in the likeness of a human mind.”
In this era of artificial intelligence, it is becoming urgent to distinguish human art from what machines produce. There is an ontological difference, even before an aesthetic one, between art and what a machine can generate through statistical calculation based on millions of images created by others. Algorithms lack the spark of humanity. For this reason, the Church wishes to renew an alliance with artists and cultural institutions to safeguard our humanity.
anyone familiar with tech knows this clip is a major bull signal for anthropic IPO
@sama knows what he has to do now but doesn’t know if he has the strength to do it
being edged to death with @cognition and @FactoryAI beef. POST THE RECEIPTS and get sloppy or shut up
most frustrating part of @X is opening a great post to read the replies and scrolling through dozens of *glaringly obvious* ai-gen slop comments
