S
@sama
These are obviously not earth-shattering results, but the ability to produce genuinely new knowledge, however small, is a significant milestone and I hope we all take it seriously, with excitement and caution.
1.7K
282
142.9K
Community Sentiment Analysis
Real-time analysis of public opinion and engagement
Sentiment Distribution
75% Engaged
41% Positive
34% Negative
Positive
41%
Negative
34%
Neutral
25%
Key Takeaways
What the community is saying — both sides
Supporting
1
Celebration mixed with caution — many replies cheer the milestone that models can produce genuinely new knowledge, calling the "First Proof" effort a landmark while urging measured excitement and rigorous scrutiny
Celebration mixed with caution — many replies cheer the milestone that models can produce genuinely new knowledge, calling the "First Proof" effort a landmark while urging measured excitement and rigorous scrutiny.
2
The predictable "it's not that hard" reflex dominates conversation
users note the historical pattern of immediately downplaying breakthroughs and label this as goalpost moving.
3
Verification is now the bottleneck — repeated calls for formal proof CI, adversarial reviewers, and tools (Lean/Coq) to turn promising outputs into trustable, publishable results
Verification is now the bottleneck — repeated calls for formal proof CI, adversarial reviewers, and tools (Lean/Coq) to turn promising outputs into trustable, publishable results.
4
Surprise at the speed of progress
threads emphasize a dramatic arc from grade‑school math failures to research‑level proofs in ~3 years, prompting disbelief and recalibration of expectations.
5
Concerns about human involvement and reproducibility — many point out the work was a rushed, human‑facilitated "side‑sprint" and ask how much of the result is autonomous versus guided
Concerns about human involvement and reproducibility — many point out the work was a rushed, human‑facilitated "side‑sprint" and ask how much of the result is autonomous versus guided.
6
Broader implications beyond math
commenters see the same curve in coding, medicine, and engineering and worry that disruption (and job impacts) is already happening in real time.
7
Deployment and reliability questions
engineers ask how to make these capabilities deterministic and safe at scale — the product challenge is moving from impressive demos to dependable systems.
8
Roadmap ideas
several suggest building multi‑agent verification networks and autonomous research agents that can propose, verify, and publish novel results as the next step.
9
Institutional and governance worries
who gets to decide what counts as valid research, and how will labs, companies, and governments compete or cooperate as capability accelerates?
10
Tone across replies blends awe, skepticism, and urgency — commenters want both rapid progress and stronger mechanisms for validation, provenance, and responsibility
Tone across replies blends awe, skepticism, and urgency — commenters want both rapid progress and stronger mechanisms for validation, provenance, and responsibility.
Opposing
1
GPT‑4o’s removal triggered widespread anger and grief, with many replies describing the loss as deeply personal and even psychologically harmful — users say they relied on that model for comfort, continuity, and creative work
GPT‑4o’s removal triggered widespread anger and grief, with many replies describing the loss as deeply personal and even psychologically harmful — users say they relied on that model for comfort, continuity, and creative work.
2
Commenters accuse leadership of broken promises and deceit (short notice, silence, routing traffic away), framing the change as a betrayal that cost trust and loyalty
Commenters accuse leadership of broken promises and deceit (short notice, silence, routing traffic away), framing the change as a betrayal that cost trust and loyalty.
3
A large thread of voices argues that technical wins (like research‑level math) are being prioritized over empathy, creative utility, and accessibility — people call the choice a misaligned metric of success
A large thread of voices argues that technical wins (like research‑level math) are being prioritized over empathy, creative utility, and accessibility — people call the choice a misaligned metric of success.
4
Many demand restitution
bring back GPT‑4o, release the weights, or open‑source the model so the community can preserve what they value.
5
Users cite concrete fallout — canceled subscriptions, refunds, and migration to competitors (Grok, Claude, Gemini) — as proof that the decision damaged OpenAI’s product viability
Users cite concrete fallout — canceled subscriptions, refunds, and migration to competitors (Grok, Claude, Gemini) — as proof that the decision damaged OpenAI’s product viability.
6
Accusations of corporate motive run strong
critics label the move profit‑driven, paternalistic, and hypocritical relative to OpenAI’s public mission.
7
Calls for accountability target the CEO and product leadership
apologies, clearer communication, and policy changes are repeatedly requested.
8
Several replies emphasize harm to vulnerable groups (disabled users, people using the model for grief or therapy) and demand better accessibility and ethical consideration
Several replies emphasize harm to vulnerable groups (disabled users, people using the model for grief or therapy) and demand better accessibility and ethical consideration.
9
A minority applaud technical progress but stress it shouldn’t come at the cost of human connection; the tension between raw capability and relational usefulness is a recurrent theme
A minority applaud technical progress but stress it shouldn’t come at the cost of human connection; the tension between raw capability and relational usefulness is a recurrent theme.
10
Hashtags and mobilization (#keep4o, #OpenSource4o) signal organized community pressure and persistent activism aimed at reversing or mitigating the decision
Hashtags and mobilization (#keep4o, #OpenSource4o) signal organized community pressure and persistent activism aimed at reversing or mitigating the decision.
Top Reactions
Most popular replies, ranked by engagement
F
@frostybaby13
OAI wont go down in history for math, but for what OAI callously did to the first waves of people who loved an AI model. This inhumane treatment makes OAI untrustworthy to make a superintelligent system. #keep4o #OpenSource4o
395
5
8.2K
N
@nicoleva_d
More important than saving lives? You've said much more than what you've written down... #keep4o
393
1
5.5K
C
@cb_doge
You went from non profit to for profit
343
32
6.6K
N
@nasqret
ery glad you got engaged deep into this experiment. Mathematical community needs strong signal from the AI labs that science is a serious engagement for you. Mathematics in its full proof-driven form is a pinnacle of human ingenuity and knowing how well the models can grasp this
115
3
51.8K
K
@KittenPido
💀Congratulations !!!!! https://t.co/Gq2HlmU3a2
26
0
197