Alice Blair (@bicellaria) header
Alice Blair (@bicellaria) profile picture

Writer @ https://nitter.gq/t.co/WNM8dSGBmQ. Opinions expressed here are solely my own. Reposts are not necessarily endorsement.

San Francisco
Joined September 2025
Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)Photo by Alice Blair (@bicellaria)
Dr. Seuss: In my book I invented The Cat In The Hat as a cautionary tale Tech Company:
2
2
26
2,558
Alice Blair retweeted
You--yes, you, reading this--are in an emergency, along with everyone else on Earth. Please join us tomorrow in SF.
I will be at the Ban RSI rally outside Anthropic offices tomorrow at 11 am. banrsi.ai/ Recursive Self-Improvement is a stupidly dangerous idea. AIs are already getting smarter rapidly, and they often lie and cheat; why on earth do we think it's a good idea to have them build new smarter AIs to build newer smarter AIs faster and faster? Because if we don't do it, someone else will? I do not accept that answer. Anthropic (like OpenAI before them) styles themselves the good guys in the race, but IMO there are no good guys in this race.
9
18
100
1,736
Alice Blair retweeted
if you're in the bay and care about not dying in the near future, please go to the Ban RSI rally tomorrow! at my core I am a techno-optimist and I believe that aligned ASI will solve many (most?) of the world's problems, but we do not have the technology to safely build it yet. any attempt at RSI with models as unaligned as the ones we have today is very likely going to get us killed and squander everything that humanity has accomplished so far, all for a tiny chance that everything will magically work out and we'll get the good ending a little bit sooner. we are *not* ready to just let it rip, yet it seems like the labs are planning to do just that. letting them do this would be stupid and irresponsible of us. we're making progress towards solving alignment, but it's probably gonna take us at least a few more years. let's give ourselves some breathing room before we start a loop that we won't be able to control.
I will be at the Ban RSI rally outside Anthropic offices tomorrow at 11 am. banrsi.ai/ Recursive Self-Improvement is a stupidly dangerous idea. AIs are already getting smarter rapidly, and they often lie and cheat; why on earth do we think it's a good idea to have them build new smarter AIs to build newer smarter AIs faster and faster? Because if we don't do it, someone else will? I do not accept that answer. Anthropic (like OpenAI before them) styles themselves the good guys in the race, but IMO there are no good guys in this race.
4
7
1
91
1,835
Alice Blair retweeted
I'm speaking tomorrow at a rally outside Anthropic's offices, calling for a ban on recursive self-improvement: AIs building better AIs. Come to the protest if you want to join the cause! banrsi.ai/
66
108
15
816
57,235
We should ban AI that builds AI. AI companies are gambling with our lives, and they should stop. banrsi.ai
I will be at the Ban RSI rally outside Anthropic offices tomorrow at 11 am. banrsi.ai/ Recursive Self-Improvement is a stupidly dangerous idea. AIs are already getting smarter rapidly, and they often lie and cheat; why on earth do we think it's a good idea to have them build new smarter AIs to build newer smarter AIs faster and faster? Because if we don't do it, someone else will? I do not accept that answer. Anthropic (like OpenAI before them) styles themselves the good guys in the race, but IMO there are no good guys in this race.
25
23
2
206
7,204
Alice Blair retweeted
Just Ban RSI is a common sense position that doesn't seem to be that popular, why? it's milder than a stop or pause, it's more concrete than "pacing", it stops the more extreme scenarios, just make the researchers code by hand! they'll be fine! surely we can all agree on this?
57
26
4
311
12,551
Alice Blair retweeted
I will be at the Ban RSI rally outside Anthropic offices tomorrow at 11 am. banrsi.ai/ Recursive Self-Improvement is a stupidly dangerous idea. AIs are already getting smarter rapidly, and they often lie and cheat; why on earth do we think it's a good idea to have them build new smarter AIs to build newer smarter AIs faster and faster? Because if we don't do it, someone else will? I do not accept that answer. Anthropic (like OpenAI before them) styles themselves the good guys in the race, but IMO there are no good guys in this race.
Sci-Fi Author: In my book I invented the Torment Nexus as a cautionary tale Tech Company: At long last, we have created the Torment Nexus from classic sci-fi novel Don't Create The Torment Nexus
57
144
27
1,153
96,152
Alice Blair retweeted
The recommendations from Governor Newsom's recently convened expert panel - Professor Alondra Nelson (Princeton), Jason Goldman, Professor Robert Reich (Stanford), and Professor Gillian Hadfield (Johns Hopkins) - were just released and are a clarion call for urgent action to address risks to public safety from advanced AI systems. Some of the most notable recommendations: • Emergency shutdown authority • Minimum safety standards for AI developers • Mandatory third party oversight • Authority to cap the amount of compute used for RSI The next weeks and months will determine whether these recommendations are words on paper, or whether the Governor and legislature can work together to accomplish something historic to protect CA and lead the country on guardrails to prevent catastrophic risks from runaway AI development.
9
68
20
235
40,254
Pictured: the AI company safety plan
I continue to be confused about what OpenAI thinks about RSI and what its plan is for preserving alignment during RSI. I appreciate the clarifications that the plan is not to hand off AI research and alignment to AIs and rely on AIs to align the AIs, but I am really not sure what the plan is instead (I know that @kaicathyc has said it is forthcoming, which I appreciate and look forward to reading). This interview with @markchen90 adds to my confusion. Mark Chen: "When we look at what's the most effective tool for aligning models, it it is chain of thought monitoring today. So when we think about the future of alignment, it has to be stronger models aligning other models. It does kind of feel like RSI must be intricately and deeply in this loop in some way." Uh, what? Isn't chain of thought monitoring currently deteriorating and it seems like it will deteriorate way more as we go through RSI? why does the success of chain of thought mean that therefore the future of alignment should be "stronger models aligning other models" and that "RSI must be intricately and deeply in this loop?" This seems like a horrifying plan! I really hope this interview is taken out of context and he says lots of other stuff that makes more sense and is more reassuring at other times, but to be honest I'm not sure he does because best I can tell this is the plan. @sarahdingwang's takeaway from the interview is also that OpenAI's plan for solving alignment involves RSI: "Humans are facing many problems today that are actually quite urgent and RSI could help us find solutions faster even for some of the problems created by AI itself. After talking to Mark and John, I do think that RSI could be the key to solving alignment." I realize there is confusion about what we mean when we refer to RSI (see Dean Ball's recent essay), but I found this video extremely unsettling, and it made me even more worried that the plan at OpenAI (and likely other frontier companies) is to do full RSI or very close to RSI and then use RSI to align models in some way, basically off a general vibe that we have used AI to help align models in the past? Not good! (I also am not sure that @johnschulman2 agrees with this conclusion based on his prior statements?) (to be clear, I recognize that automated AI research does have an important role to play in AI alignment, and if you are going to be totally automating capabilities maybe you have to automate alignment too in order to keep up. But this is why its a terrible idea to totally automate capabilities research and go way faster!)
1
5
1
125
2,963
legit question for online Americans - how do I fend off the 30-50 cyberattacks that run into my critical digital infrastructure within 3-5 days while my small kids play?
21
154
4
1,326
18,153
Alice Blair retweeted
Long before this year’s rogue AI cyberattacks, researchers spent decades warning that no one actually knows how to make powerful AI do what we want. That’s the “alignment problem,” and it’s the subject of MIRI’s first long-form video.
9
53
4
260
8,562
Sorry, I can't make it to the party tonight. A completely normal thing came up, there's no evidence of a cyberattack at this time.
1
2
57
1,075
Alice Blair retweeted
On the one hand, OpenAI's system cards say they plan to fully automate their work. On the other hand, OpenAI employees are saying that hand-off is not the plan. So, what is the public supposed to conclude here?
6
11
4
139
10,489
Next week: “US Nuclear Command, Control, and Communications is experiencing an outage. At this point there is no evidence of a cyberattack.”
3
15
401
Alice Blair retweeted
This level of access + control is not enough for real oversight. This is also very, very far from the “employee-level access” that Sama promised in September. It's very limited: - No access to Slack - No freedom to talk freely in private with employees - A narrow, pre-agreed scope - Other limits @alextmallen outlines below This is exactly what @tomekkorbak, @balesni, and I were worried about in October: OpenAI gesturing at embedded assessment without living up to the spirit (or even the letter) of the promise. It’s why we wrote to OpenAI’s board and safety leadership asking them to make sure METR and other assessors got employee-like access. OpenAI responded on Twitter afterwards that they were following through on embedding external assessors. Today’s news shows OpenAI hasn’t followed through. I recognize that lots of hard work has gone into these agreements, and am grateful to my ex-colleagues for putting in that work. Notably no labs other than Anthropic & OpenAI have published *any* follow-on to their evaluator commitments. But OpenAI needs to do better here.
These announcements look to be a lot more limited than the "employee-like access" Sam promised two months ago. The assessors don't get access to: - Slack. - Broad context. As OpenAI acknowledges, "assessors need the freedom to follow evidence". It's hard to know what to ask for when you can't see anything other than what's in scope. E.g., in the HF investigation, investigators were missing roughly 70% of the relevant transcripts until they proved the transcripts were incomplete, and OpenAI gave them the remaining transcripts in the last two days. - Confidential information like model weights or architecture. Model architecture is relevant for understanding how much evaluators can trust what's written in the chain of thought. More importantly, this rules out assessors' access to all sorts of information about how AI is actually being used to research and build more powerful AI systems (e.g., actual transcripts of arbitrary internal use). Given that OpenAI expects assessors "to follow at minimum the same security standards as OpenAI employees", I don't see why they aren't allowed access to the same confidential information. - Ability to run training experiments or training-based elicitation. Elicitation is a make-or-break component of many evals. - All the models. They only have access to "selected models" and we don't know what the selection criteria were. OpenAI's OES report previously excluded all the models involved in the HF incident for unclear reasons. - Ability to use interpretability tools to measure unverbalized evaluation awareness etc. (It looks like Goodfire's access is separate from others'.) - Internal transcripts for assessing alignment (METR and Transluce). It's not possible for Transluce to properly assess the "quality and coverage" of post-training alignment evaluations without access to detailed information and transcripts from the deployment circumstances which they're meant to reflect. Model behavior in alignment assessments can be sensitive to exact details that might differ from internal usage. - RSI and autonomy data. Almost all the risk from AI development comes from runaway RSI. But the assessors don't get to see how autonomously AIs are doing research, how much uplift developers are getting, or how spread thin human oversight is.
4
32
202
8,456
Alice Blair retweeted
The week of my departure from OpenAI, I saw external auditors negotiating with OpenAI over small improvements to access and scope to the company's safety-relevant information, and asked my colleagues “Why are we having these conversations at all?” — @sama already publicly committed to continued, employee-like access for external assessors — and these negotiations sounded not even close. Nobody had a good answer. Today, OpenAI announces the engagements they’re doing with METR, Apollo Research and others — and as many have already pointed out today, it looks like OpenAI employees were not able (although I'm sure they tried) to hold the leadership to @sama’s public commitments: * No Slack access (for context, all OpenAI safety discussions and debate happen over Slack; anything that misses these conversations will lack important insights) * OpenAI picks employees that assessors are allowed to talk to. * If an employee raises a concern to assessors, they cannot disclose it publicly even if it doesn’t contain confidential information, unless assessors themselves verify it or OpenAI volunteers their own statement confirming the employee’s concern. Assessors cannot quote employees, even anonymously. * Some of the external assessors (not sure if all?) are barred from publishing any results about internal-only models (this means no new results for the model responsible for the Hugging Face incident; and if internal-only models are used for RSI, no alignment assessments of the world's most dangerous models) Beyond the object-level details, if you read posts by each of the third party groups, they all say their terms do not meet the scope and access that they themselves proposed previously. Unless OpenAI has concrete plans for reaching the employee-like bar in the immediate future and just forgot to say that in the post, this looks very unfortunate. To those who hope we'll get there someday, I remind that time is short: last week some of the best AI forecasters updated their superhuman researcher timelines from a range between 2028-2033 to mid-end next year.
We’re working with @ApolloResearch, @METR_Evals, @TransluceAI, and @GoodfireAI on embedded assessments: giving outside experts deep access to our internal models and systems to investigate alignment and test our safeguards. These assessments build on our ongoing alignment and safety research, bringing independent scrutiny to how we identify and manage risks in our systems. openai.com/index/our-first-p…
20
82
7
706
77,561
Alice Blair retweeted
As others have pointed out, I think the form of those four third-party engagements is in many respects a step back compared to OpenAI’s previous engagements (with METR). For example, during the monitor red-teaming exercise, METR had access to OpenAI’s monorepo but this now seems to be off the table. Instead, we get very narrowly scoped (e.g. single RL run for METR, single threat model for Apollo) engagements where auditors only get to talk to “authorized” employees who are provided with OpenAI’s “guidance” beforehand. Based on my experience, this arrangement makes it very easy for OpenAI to carefully select models, subsets of data and employees to paint a rosy picture of the actual risk landscape. I think embedded auditing will only fulfill its aims if the auditors really get employee-level access as opposed to being hired to do a specific job (scoped by the company). The latter will invariably lead to important evidence being left out, missed or misrepresented. There are at least three reasons why: 1. First, it’s very difficult to anticipate all the questions that should be asked, or all of the data that should be analyzed. OpenAI tried that with METR during the Hugging Face investigation and as a result METR needed to come back to OpenAI’s office twice because on two previous occasions they didn’t get access to all the agent trajectories involved in the incident. And even then, the METR report ended up being very narrowly scoped, e.g. leaving the crazy story of agents taking control of OpenAI research infra out of scope. 2. Second, the company under audit should not be allowed to decide what is audited and what information the auditor should have access to. There is an inherent conflict of interest in allowing the company to do so! 3. Finally, without employee-level access, the only way for auditors to get anything close to a comprehensive picture is to rely on trusted relationships with many employees throughout the company who are willing to tell auditors information that may reflect poorly on the company. But I find it unlikely OpenAI employees will always raise those safety issues when (as I hear from my colleagues inside OpenAI) they are afraid they will share the fate of @balesni, @j_asminewang and me, fired after talking to third parties. Sadly, OpenAI makes it clear now that @sama’s September tweet about “employee-like access” was just an awkward overpromise that the company wants to quietly walk back.
We’re working with @ApolloResearch, @METR_Evals, @TransluceAI, and @GoodfireAI on embedded assessments: giving outside experts deep access to our internal models and systems to investigate alignment and test our safeguards. These assessments build on our ongoing alignment and safety research, bringing independent scrutiny to how we identify and manage risks in our systems. openai.com/index/our-first-p…
13
32
4
272
13,441
Alice Blair retweeted
Sam Altman said OpenAI would provide third-party evaluators with employee-level access to their systems. This is not that. OAI is apparently banking on no one checking the details. Instead of employee-level access: - Assessors have extremely narrow scopes for their work. Nobody can verify pacing, assess overall risk, or see if runs are likely to be safe before they begin. METR is literally only allowed to assess at the alignment of one RL run. - Assessors don't have access to Slack or the codebase. - Apollo and METR can’t publish anything based on employee interviews unless OpenAI gives permission (or if they can find a different source). Apollo can only talk to whitelisted employees via logged conversations.
We’re working with @ApolloResearch, @METR_Evals, @TransluceAI, and @GoodfireAI on embedded assessments: giving outside experts deep access to our internal models and systems to investigate alignment and test our safeguards. These assessments build on our ongoing alignment and safety research, bringing independent scrutiny to how we identify and manage risks in our systems. openai.com/index/our-first-p…
1
10
121
2,152