Timo Koriseva's Avatar

Timo Koriseva

@koristi.bsky.social

Watching anime, F1 and the news. Performing improv and standup. Raising a tiny human. Trying his best to keep things working.

56 Followers  |  52 Following  |  55 Posts  |  Joined: 25.10.2023
Posts Following

Posts by Timo Koriseva (@koristi.bsky.social)

Post image

Talviterveisiä perheestä! Kuva otettu jo jonkin aikaa sitten, kun oli vielä pakkasia.

01.03.2026 12:19 — 👍 6    🔁 0    💬 0    📌 0
Preview
Tampereen roolipeli- ja animetapahtuma Tracon on valittu vuoden 2025 kulttuuriteoksi [Tampereen kaupunki] Tampereen kaupungin kulttuuri- ja vapaa-aikalautakunta palkitsee vuoden kulttuuriteon tunnustuspalkinnolla. Tämän vuoden syksyllä 21. kertaa järjestettävä Tracon on kasvanut pienestä

tampereen kaupunki on valinnut traconin vuoden 2025 kulttuuriteoksi

www.tampere.fi/ajankohtaist...

#tracon

19.02.2026 14:36 — 👍 55    🔁 16    💬 0    📌 0
Post image

Tää

23.01.2026 09:35 — 👍 26    🔁 3    💬 1    📌 0

Jos fiktio on kohdemaassa tuotettua, niin silloinhan se on kirjaimellisesti kyseisen maan kulttuuria, jota faktateokset eivät ole, vaikka kulttuurista kertoisivatkin.

31.12.2025 07:00 — 👍 1    🔁 0    💬 1    📌 0

Where can I watch it?

25.12.2025 09:32 — 👍 1    🔁 0    💬 1    📌 0

Eikö kuva toisaalta implikoi, että Cruise-aneki nousee haudasta? 🤔🤔🤔

01.11.2025 21:11 — 👍 5    🔁 0    💬 1    📌 0

Toki Tampereelle kannattaa tulla Traconin aikaan vaikka kaljalle!

06.09.2025 20:39 — 👍 0    🔁 0    💬 0    📌 0

"Välttämätöntä" "Pakko" "Ei vaihtoehtoja" ovat noin muutenkin nykyhallituksen suosikkiperusteluja.

22.08.2025 17:39 — 👍 0    🔁 0    💬 0    📌 0
Post image

Auringon vinkit. (2021)
#faksiarkistoista

02.08.2025 10:53 — 👍 23    🔁 2    💬 0    📌 0
Post image 11.07.2025 10:33 — 👍 16917    🔁 4761    💬 208    📌 218
The system prompt defines the character Claude is supposed to play: an intelligent assistant with “depth and wisdom”, possessing its own thoughts and point of view, capable of emotions like interest and enjoyment. It’s told explicitly not to claim that it lacks sentience or subjective experience — rather, to treat the question of its own consciousness as intriguingly unknowable.

How does Claude understand, purely on a language level, what any of this means? How does it know what a friendly, possibly-sentient AI is, or how such an entity is supposed to act? Like any LLM, it draws on what exists in its training data. And since an “intelligent assistant” is something which, until very recently, was only a speculative concept, Claude’s behavior — and the behavior of every AI assistant — is constructed largely from what exists in fiction.

The system prompt defines the character Claude is supposed to play: an intelligent assistant with “depth and wisdom”, possessing its own thoughts and point of view, capable of emotions like interest and enjoyment. It’s told explicitly not to claim that it lacks sentience or subjective experience — rather, to treat the question of its own consciousness as intriguingly unknowable. How does Claude understand, purely on a language level, what any of this means? How does it know what a friendly, possibly-sentient AI is, or how such an entity is supposed to act? Like any LLM, it draws on what exists in its training data. And since an “intelligent assistant” is something which, until very recently, was only a speculative concept, Claude’s behavior — and the behavior of every AI assistant — is constructed largely from what exists in fiction.

A chatbot’s system prompt is the invisible beginning of every conversation it has with a user. (The final line of Claude’s prompt is “Claude is now being connected with a person.”) When the user types a question or instruction, the text they enter is appended to the end of the system prompt, and the LLM is then made to generate more words — a few sentences or paragraphs that seem appropriate, on some deep-probabilistic level, as a continuation of the existing text — which are presented to the user as a response.

To be totally clear about what’s happening here: the language model is doing next-token prediction on a document whose first several thousand words establish everything that follows as a conversation between a human and a highly advanced artificial mind — a type of mind that doesn’t really exist, but that the model is designed to roleplay. When a person uses Claude or ChatGPT, they’re meant to believe they’re interacting with a helpful AI assistant. What they’re actually doing is collaborating with a language model on writing a work of science fiction about a human talking to an AI.

A chatbot’s system prompt is the invisible beginning of every conversation it has with a user. (The final line of Claude’s prompt is “Claude is now being connected with a person.”) When the user types a question or instruction, the text they enter is appended to the end of the system prompt, and the LLM is then made to generate more words — a few sentences or paragraphs that seem appropriate, on some deep-probabilistic level, as a continuation of the existing text — which are presented to the user as a response. To be totally clear about what’s happening here: the language model is doing next-token prediction on a document whose first several thousand words establish everything that follows as a conversation between a human and a highly advanced artificial mind — a type of mind that doesn’t really exist, but that the model is designed to roleplay. When a person uses Claude or ChatGPT, they’re meant to believe they’re interacting with a helpful AI assistant. What they’re actually doing is collaborating with a language model on writing a work of science fiction about a human talking to an AI.

Wrote about what's really going on when you talk to an AI assistant

03.07.2025 20:13 — 👍 461    🔁 131    💬 7    📌 21
Post image 04.05.2025 19:41 — 👍 4510    🔁 1993    💬 37    📌 78
Post image Post image

Desuconeissa useasti vieraillut Rauha-opaskoira on jäämässä eläkkeelle.

Desucon kiittää Rauhaa hyvästä työstä! 💚

15.06.2025 12:16 — 👍 185    🔁 34    💬 2    📌 0
Post image 30.04.2025 20:41 — 👍 27    🔁 14    💬 0    📌 0

Ah, mul on tämmönen messy boi kans.

20.04.2025 20:19 — 👍 0    🔁 0    💬 0    📌 0

Kannustitko hankkimaan sitten joskus sellaisen työn, että siellä ois kiva olla vaikkei niin maksettaiskaan?

19.03.2025 07:11 — 👍 1    🔁 0    💬 1    📌 0

Katsottiin KAJ:n 10v-juhlakonsertti YLE Areenasta, joten pidän itseäni nyt #KAJ -asiantuntijana.

15.03.2025 12:24 — 👍 0    🔁 0    💬 0    📌 0
Preview
a little girl in a white karate uniform with the word eyaaa written on the bottom ALT: a little girl in a white karate uniform with the word eyaaa written on the bottom

Koska he ovat liian kiinnostuneita KARATESTA. T: kuukausi takana peruskurssia

14.02.2025 22:21 — 👍 2    🔁 0    💬 1    📌 0

Niin, sitten voi kysyä, että mistä nämä nuoret ovat oppineet tuon palvelualan väheksynnän?

10.02.2025 21:26 — 👍 0    🔁 0    💬 1    📌 0

Nk. valistunut diktatuuri on ilmeisesti teoriassa kaikista paras ;)

09.02.2025 15:29 — 👍 1    🔁 0    💬 1    📌 0

Jep, hitaus ja politikointi on se hinta, jonka maksamme siitä, että yhteiskunta pysyy edes jotakuinkin vakaana. Sitten kun ihmiset alkavat toimia muuten kuin itsekkäästi ja järjettömästi, niin voimme miettiä joitain muita vaihtoehtoja hallintomuodoille.

09.02.2025 11:44 — 👍 7    🔁 1    💬 1    📌 0
DesuTV Frostbite 2025 Kooste
YouTube video by Desucon DesuTV Frostbite 2025 Kooste

tehtiin taas tämmönen tapahtuma

#desucon #desuconfrostbite

26.01.2025 15:11 — 👍 29    🔁 8    💬 0    📌 0

If its between +1, 0, or -1 Celcius, then absolutely yes.

26.12.2024 11:24 — 👍 2    🔁 0    💬 0    📌 0

Niin se aika rientää. Ei ehdi paljoa someen postailla ja jo oli prinsessan 1v-syntymäpäivät!

13.10.2024 19:56 — 👍 7    🔁 0    💬 0    📌 0
Post image

Sarjassamme...

26.09.2024 09:24 — 👍 0    🔁 0    💬 0    📌 0

Mulle siellä näkyy ihan liikaa ihmisten tosi henkilökohtaisia päivityksiä a'la Facebook, mutta julkinen. En tiedä onko se sit engagement farmingia vai mitä, mut kun jengi avautuu vaikka niiden lasten sairauksista niin on melko erikoista.

24.09.2024 20:59 — 👍 0    🔁 0    💬 1    📌 0
A virtual certificate with text "Celebrating 10M users on Bluesky, #1,713,363, Timo Koriseva ‪@koristi.bsky.social‬, joined on Oct 25, 2023"

A virtual certificate with text "Celebrating 10M users on Bluesky, #1,713,363, Timo Koriseva ‪@koristi.bsky.social‬, joined on Oct 25, 2023"

Bluesky now has over 10 million users, and I was #1,713,363!

20.09.2024 17:45 — 👍 1    🔁 0    💬 0    📌 0

Hmm, missähän tällaisista weebirundeista kenties ilmoitellaan?

03.09.2024 16:54 — 👍 0    🔁 0    💬 0    📌 0

Use case ja syy käyttää rahaa luujohdekuulokkeisiin: Voi kuunnella asioita hiljaa, mutta kuulla myös jos prinsessalla on jokin huonosti makuukamarissaan.

12.07.2024 18:09 — 👍 0    🔁 0    💬 0    📌 0

Onnittelut KcS ry:n uudelle hallitukselle ja uudelle puheenjohtajalle @jarnz.bsky.social ! #desucon

06.07.2024 17:18 — 👍 4    🔁 0    💬 1    📌 0