dothermit

[dothermit @DottedHermit]

I was able to get about 10 mins of usage (time model spent working) with astra-6 low in plus plan. 5 hour limit hit by this time. I sent 7 messages. It did optimize my project and created plan for more optimization within this session. So, was pretty useful wish had more usage.

[♡Gσ∂∂єѕѕ ℓуѕѕα♡ fιи∂σм @juicylyssa20]

Without me you would be absolutely miserable. Remember that next time you try running🤣

[LuckyHermit @dotherm123]

As a sub if u feel like you have to quit findom, then realize quickky that its a wrong feeling

findom

[LuckyHermit @dotherm123]

need drain training

findom paypig

[dothermit @DottedHermit]

@fchollet A very high working memory if possible will be the biggest advantage. I think us humans if we had a much higher working memory would be able to solve a lot more problems.

[primordial intelligence @primal_brainer]

hi twitter,

i am looking for a 2BHK in Whitefield / Brookefield in a gated society.

budget: ₹45k/month (ince maintenance).

if you know of any good leads, I'd really appreciate it.

repost for 1000 aura points :)

[dothermit @DottedHermit]

Narrator : They did not

[Fabrizio Romano @FabrizioRomano]

🚨🏴󠁧󠁢󠁥󠁮󠁧󠁿 Thomas Tuchel: “We will find a way to stop Leo Messi”, told Telegraph.

[dothermit @DottedHermit]

@paraschopra But you can backtrack in your speech and agents understand that That also gives context to agents that u considered that idea and backtracked. You can speak for 2-3 minutes describing and it doesnt have to be in order or shape. Agents just get full context and make sense of it.

[dothermit @DottedHermit]

@DeryckOE @thsottiaux What about 5.5 in low thinking mode? That should cost less.

[dothermit @DottedHermit]

@0xglitchbyte @ForrestPKnight They are doing that and the current stable bun version is still zig based. Once they clean up only they plan to make the rust version the stable.

[dothermit @DottedHermit]

@ZoomerRepent @perrymetzger You mean a human solved it and attributed it to GPT? And that human happened to be an openai employee? Or some human solved and contacted openai and they paid huge sums of money to get it attributed to open ai? All of these seem way less probable than the straightforward scenario

[dothermit @DottedHermit]

@ExpressoMarc @ihteshamali @big_killa_ The brainstem that starts this dream doesn't know that the person is blind. It's a general mechanism in all humans even though it doesn't really help blind people, it still fires.

[dothermit @DottedHermit]

@joodalooped that quote makes a lot of sense not sure what's wrong

[dothermit @DottedHermit]

@_antmad @iam_polymath @sledge_baba they can keep track of which side the stick was when they slept and when they wake up. circle not needed again

[dothermit @DottedHermit]

@etarasietliet @Pask218337 @WelshUpasaka @Babygravy9 I agree with the other guy. We have strong in-group bias. The idea that female homo sapiens might have preferred male neaderthals is just feel good thinking. Just think of even within same species how we have so many racial and other issues due to small differences.

[dothermit @DottedHermit]

@jdegoes i have had success by adding just a skill which contains a concise doc of the language and it just works well. A new language doesnt have to be in the training data.

[dothermit @DottedHermit]

@MANSAFAB Isnt it 1.6g per lean mass weight. So, not 60kg it will be like 1.6 * 48, if around 20% body fat.

[dothermit @DottedHermit]

@amritwt AI always has a way of suprising like this. It happened with art and with math and coding. So, i think now when we least expect it robots will come capable of doing manual work and we will be surprised again

[dothermit @DottedHermit]

@code_kartik the real world lags by atleast 6 months. So, you will hear a lot soon.

[dothermit @DottedHermit]

@Blueelectron4 @redaction thats the outdated way to look at it. Once a general model is trained it can learn new skills by just having a documentation or guide to it. You dont need to train it with the same data.