[ExI] OpenAI Says It Will Not Release Newest A.I. Model Over Safety Concerns

[Jason Resch]

On Tue, Sep 29, 2026, 3:13 PM John Clark <johnkclark at gmail.com> wrote:

> On Tue, Sep 29, 2026 at 12:35 PM Jason Resch via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
> * >>>  I think it's just their marketing strategy.*
>>>>
>>>
>>> *>> Could you clarify what you mean by that? *
>>>
>>
>> *> It excites their investors to hear "Our AI is so good, we can't even
>> release it."*
>>
>
> *A senior researcher from open AI, Jacob Coxon recently resigned, and said
> goodbye to tens of millions and perhaps hundreds of millions in stock
> options, because he said the company was not paying enough attention to
> safety. Do you think he did this for some convoluted Machiavellian reason?*
>


People can be moral and stand on principles. Profit seeking companies
usually can't; they get supplanted by their less scrupulous competition.


>
> * > And it is also a convenient catch-all excuse when they run behind
>> their planned release schedule.*
>
>
> *Given the huge advance in AI capabilities made in just the last three
> months, I don't think any AI company needs to make up excuses for slow
> progress.  *
>


They have an IPO coming up. Look how much free press this announcement got.
And of course for their potential customers: everyone is drawn to want to
use the AI that's considered "too dangerous" to release to the general
public--who cant be trusted with such power. (but for just $19.99/mo you
can).




>
>
>> *>>  Regardless of what the motivations of AI companies are for saying
>>> what they're saying, do you think AI is unlikely to produce changes to our
>>> civilization of unprecedented size?  *
>>>
>>
>> *> Where did it appear I said that?*
>>
>
> *I wasn't accusing you of anything, I just didn't understand your
> position, and I still don't.  *
>


It's: OpenAI has been saying the same thing for the past 6 years, but it
has never stopped them, so I don't expect them to hold back 6.1 either. And
I add this hypothesis: their fear mongering also suits the company's goals
for boosting valuation and attracting hype.




>
>> * > they're are doing a disservice to those trying to get things done for
>> AI safety.*
>>
>
> *Open AI said they decided not to release their most advanced AI model and
> renounce billions of dollars of revenue in the proces, (a rather poor
> marketing strategy in my humble opinion) because they didn't want a repeat
> of the hugging face incident, or something much worse. And they told
> everybody why they were not releasing it.*
>

Just like they've said since GPT 2.0. Have you forgotten they said the same
thing about Astra 6.0 and Fable 5 a few months ago? I have access to them
both of them today.

* But you are unhappy with OpenAI's actions.*
>


Please don't attribute states of mind to me. I've neither said I support
nor that I do not. I'm just questioning the veracity of their promise to
not release it and their motives for saying they will not given their
consistent history of still going ahead and releasing models after claiming
it was too dangerous to release.


* So what do you think they should've done instead? Should they have
> released the model anyway? Should they have quietly killed it but kept
> top secret the reason why? I'm not trying to put words in your mouth, I
> just don't understand exactly what your complaint is.*
>

I'm making no complaint, only adding the context of their historical claims
and breaking of promises. So when they make the same claim about GPT 6.5 or
7, after 6.1 has already been released, you will remember to take that
claim with a grain of salt.

Jason


>
>
>
>
>
>>
>>>
>>>
>>>>
>>>> On Tue, Sep 29, 2026, 7:55 AM John Clark via extropy-chat <
>>>> extropy-chat at lists.extropy.org> wrote:
>>>>
>>>>> Explore this gift article from The New York Times. You can read it for
>>>>> free without a subscription.
>>>>>
>>>>> OpenAI Says It Will Not Release Newest A.I. Model Over Safety Concerns
>>>>>
>>>>> The company’s researchers raised questions about the security of the
>>>>> model, known as GPT-6.1 Astra.
>>>>>
>>>>>
>>>>> https://www.nytimes.com/2026/09/28/technology/openai-astra-safety.html?unlocked_article_code=1.E1E.s0GM.WjUmxr83iqqM&smid=em-share
>>>>>
>>>>
>>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260929/7859d161/attachment.htm>