Google's Logan Kilpatrick posts that the model is the product as replies examine models operating as components inside agent-driven systems with continual learning and user experiences
Discussions address whether full systems determine AI outcomes.
#64Nathan Lambert@NATOLAMBERTThe system is the product. Models are just one piece today, agree with Joanne.
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
@joannejang not enough people talking about this switch imo. Model + harness + tools all are core components, it's a system.
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#83rohan anil@_AROHAN_Research innovations drive model progress.
Logan Kilpatrick@OfficialLoganKThe model is the product
#83rohan anil@_AROHAN_Working at core automation felt different a day or so after Joanne joined, and its been an incredible time.
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
@eliebakouch @joannejang Many such cases
elie@eliebakouch@joannejang lol i had no idea this quote was from you originally 😮
#160Omar Khattab@LATEINTERACTION
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#244Joanne Jang@JOANNEJANGbanger
@alth0u we might be talking about different time horizons
people used to read CoTs and diffs but are doing so less as models get better and earn trust
i enjoy autopilot but the future is fsd & waymos where you don't have to think about transportation unless as a hobby
alth0u🧶@alth0ui strongly believe the complete opposite of this: - the computer should be invisible - it should be obvious there is a model doing things next to you you should feel the car hum you should be able to shift gears no skeumorphic bullshit in the age of ai anymore
learned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026!
The model should be invisible.
i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#437Xiao Ma@INFOXIAO看山还是山
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#716elie@ELIEBAKOUCH@joannejang lol i had no idea this quote was from you originally 😮
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#867Alexander Doria@DORIALEXANDERTo be honest also have been updating lately, but mostly because product is underselling it.
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#867Alexander Doria@DORIALEXANDER
Logan Kilpatrick@OfficialLoganKThe model is the product
@OfficialLoganK obviously
Logan Kilpatrick@OfficialLoganKThe model is the product
#1114alth0u🧶@ALTH0Ui strongly believe the complete opposite of this: - the computer should be invisible - it should be obvious there is a model doing things next to you
you should feel the car hum you should be able to shift gears no skeumorphic bullshit in the age of ai anymore
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#1220rohit@KRISHNANROHITmodels are the fuzzy processors inside the complex analytic systems of tomorrow
Joanne Jang@joannejanglearned this quote from 2023 is making rounds -- i actually don't think this is true anymore in 2026! The model should be invisible. i expect us to flip back to ux in the form of agent behavior + continual learning loops; and the alpha is in making models feel natural and as invisible as possible.
#1703alex peysakhovich@ALEX_PEYSthe most important lesson to internalize for any ai or ai augmented company
Logan Kilpatrick@OfficialLoganKThe model is the product
there's two competing strategies playing out:
better models vs. better products