I don't understand this, for example, what would you have done if you where Ukrainian right now ? (before 2014 arguably start of conflict and after invasion)
That is an interesting question, very far from my daily concern and brings dilemmas when I think about it. My response would probably be "I don’t know".
However Anthropic situation is very different: there’s no ongoing invasion of the USA, and they traditionally attack other countries once in a while (no judgment) so the weapons upgrade will be "useful" on the field.
It is of course possible to argue that the reason there is no ongoing invasion of the USA is because of our continued investment in technology for killing people
I am 100% against militarism and wished we didn't need any of this, but the power balance between Russia and Ukraine or even Israel and the Palestinians seem to corroborate the thesis... There likely would be no Ukraine war today if Ukraine hadn't voluntarily given up its nukes three decades ago (unproven thesis). There was one as Russia thought it could win. The ongoing (after the "peace fire") Israeli occupation and attacks of the remnants of Palestinian territory show the same. If you are the weaker party and there is a stronger party that wants what you have (or plain wants to eradicate you) then they'll do so..
> I don't understand this, for example, what would you have done if you where Ukrainian right now ? (before 2014 arguably start of conflict and after invasion)
There are a lot of well meaning people that are very anti-weapon or anti-violence under any circumstances. The problem is that when those people actually need those weapons and that violence, they are so inadequate at it that they become a liability to themselves and others.
I'm not saying I have or know of a solution, but I remember the old saying (paraphrasing) that it's better to be a warrior working a farm than a farmer working a war.
Or may be education should be more dynamic, engaging, and interactive, instead of having lowest paid teacher jobs, with overcrowded classes, heavenly focused on boring memorization (without clear purpose), and boring tests.
in modern day, at least stories I heard, criminals will be placed in temp holding sell (worst places), once max time in temp jail is expiring (legally allowed), they just add new charges, so effectively, you will be held in worst jail cell possible for weeks if not months, this is how I understood from stories of multiple life time criminals with 10+ years in jail time served.
I think there is a language issue here (doesn’t seem like English is their first language?), so I suspect the original comments intent was different than what it actually says
GPT has no sense, or care when it is wrong or right, such sense is only (arguably) driven by human through prompt interaction and throughout training of model, while humans and other animals able to update there's internal state just from single observation or interaction, and integrate future information with such single observation for very long time.
And other inputs from the environment. Then there were mechanisms which could also be neural networks that will transform this data into more digestible way for GPT and GPT was also in addition specifically trained to act based on this input.
Then it would run in cycles, where it gets this input, and it will provide output on how it plans to react to the data, maybe every 100ms.
It then could also have a storage it can use, where it can store data as part of the output to later retrieve it again.
So it would be a set of modules that is controlled and interpreted by GPT.
It could then do all of that above, no? And all of it should be just a matter of implementing. The only near time challenges may be certain types of inaccuracies and or producing tokens in some cases might take too long time to have fast reaction time.
So basically you'll try to run as frequent cycles as possible with the inputs mentioned above, other neural networks identifying the objects, in many different ways and all the context about the environment, unless a new version of GPT becomes completely multi-modal.
And you run those loops, then GPT gives output what it wishes to do, e.g. store some fact for later usage, move there, move here, etc. Or retrieve some information using embeddings then decide again, and short term memory would just be this context sized window, and if it needs more it just looks into its own memory for embeddings.
how system updates and maintains own model(s) when new information added in form of single/few observation and/or interactions, without ability to replay data, without catastrophic forgetting etc..., and importantly, how such system's model(s) grow in complexity while retaining stability and redundancies.
I tried Optery, Incogni, and long time ago OneRep, way to lazy to do it myself, don't worry they will have my info, data is already on internet.
Incogni at least in there's claim offers opt-out from private databases (no way to verify ) and some but not all public database (eg. google searches).
Optery has largest list of public databases (with most expensive subscription) out of everyone else, there's costumer service is responsive regarding failed removal.
OneRep was not bad long time ago when they run it from Belarus (I know, crazy), but they would refresh somehow search caches too (it could be ok, or make things worse), they don't seem to offer advertise this service any more.
Don't search your self only via google, for example, bing will give different results, some databases will have misspelled names (could be deliberately), so there still some work to make sure all records are removed.
At this point, this is like privacy tax that you have to budget to have at least your address on cell phone number not easily discoverable.
seems like wrong case study, cli power (at least for me) is in comments and record keeping what has been done months if not years ago, and an "easy way" to script repetitive tasks, I just don't see how same thing can be achieved with GUI:
record video and then comment it ?
have detailed logs what button, and what fields where edits/pressed ?
I'm going to guess that that database moved to RDS and, once there, was likely refactored as services were decomposed to microservices... but that's not mentioned in the article.
There is not way in hell, I will be able to work for such company more then 6 month (well may be it if it is 7 figures, but even then I don't think I last longer then one year), I am sorry who has to go through this.
reply