Petter Bae Brandtzaeg

Working at @UniOslo & @SINTEF_HCI researching the social impact of people's uses of new technologies, such as social media and conversational AI

[Petter Bae Brandtzaeg @PetterBB]

Publication alert: New Lancet Commission on liver health. My contribution examines how social media algorithms may shape food and alcohol-related choices, especially among young people. https://t.co/SBUdu2EQmj

[Petter Bae Brandtzaeg @PetterBB]

New publication on human-AI relationship - AI individualism https://t.co/jZbVkNkoyC

[Petter Bae Brandtzaeg @PetterBB]

New publication: Written about our new concept AI individualism, that try to explain how AI are changing the social structure in society. @asbjrn https://t.co/7OuBdvRRzj

[Eric Topol @EricTopol]

This week's @Nature cover highlights a report of A.I. mediated distortion
When #ChatGPT was asked to rate 40,000 résumés, it ranked the older male candidates as better quality than the younger female applicants https://t.co/qTSHdxF8Ia

[Petter Bae Brandtzaeg @PetterBB]

Å si at "byrådet begynner å få kontroll på" elsparkesykkene er ikke godt nok. Enten har man kontroll, el så har man det ikke. Med så mange alvorlige skader, er d en skandale at byen lar det fortsett @Venstre @OsloHoyre https://t.co/Uny0Yhlbab

[Petter Bae Brandtzaeg @PetterBB]

Så bra @aftenposten.no trykker kritikk av Aftenposten. Kritiserer Aftenposten i dag for å generere en KI-rapport og å likestille det med forskning https://t.co/Gxi2iBXsS4

[hardmaru @hardmaru]

Midjourney can generate pictures of Biden, Trump, Putin, the Pope, etc, but it will not generate Xi Jinping. 🤔

“Without uniform standards, individual companies are deciding what’s permissible—and, in this case, when to bow to authoritarian governments.”

https://t.co/LNRIjDR8s7

[Petter Bae Brandtzaeg @PetterBB]

Publication alert: Brandtzaeg, P. B., Skjuve, M., & Følstad, A. (2025). AI individualism: Transforming social structures in the age of social artificial intelligence. In P. Hacker (Ed.), Oxford Intersections: AI in Society (online ed.). Oxford Academic. https://t.co/3wkZ8geHqq

[Mark R. Hinkle @mrhinkle]

Generative AI is changing filmmaking—if you can describe it, AI can create it.

This video, made with MidJourney and Kling AI by Jacopo Reale, proves how realistic AI-generated cinematography has become.

[Petter Bae Brandtzaeg @PetterBB]

@realDonaldTrump approval rating sits at 44% (Reuters, 19.02.2025), while @ZelenskyyUa is at 57% (KIIS, 19.02.2025). Yet, Trump claims Zelenskyy only has 4% support. 🤔 Whose numbers are you trusting? 📊 #ApprovalRatings #FactCheck #Trump #Ukraine #UkraineRussiaWar

[Petter Bae Brandtzaeg @PetterBB]

Kronikk om ChatGPT vs Fagpersoner i på https://t.co/yftGyDc6cU https://t.co/0HfhIaKsqX

[Sam Altman @sama]

insane thing: we are currently losing money on openai pro subscriptions!

people use it much more than we expected.

[Petter Bae Brandtzaeg @PetterBB]

Good news from Meta "We're ending our third party fact checking program and moving to a Community Notes model."

[Petter Bae Brandtzaeg @PetterBB]

Digital dyneløfting. Vi feiltolker ofte det digitale. Og bedre vil det ikke bli med utbredelsen av kunstig intelligens. Det digitale er ofte fastspikret. Det lagres. Det er søkbart. Det er manipulerbart. Det er delbart og sårbart. https://t.co/lbAYe6C40A

[Bindu Reddy @bindureddy]

OpenAI has released ZERO features for developers in the last 7 days

They went from open source to closed source to no API access AND give us $200 to access our top LLM!!

EXTREMELY ALARMING!

[Deedy @deedydas]

Claude just shipped “Computer Use” which allows an LLM to control your computer and do arbitrary things like browse websites, download and run files, and more.

A slew of startups doing browser agents feel obsolete overnight.

Seriously cool stuff.

[Aurimas Griciūnas @Aurimas_Gr]

𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 𝗟𝗼𝗻𝗴 𝗧𝗲𝗿𝗺 𝗠𝗲𝗺𝗼𝗿𝘆 explained simply.

In general, the memory for an agent is something that we provide via context in the prompt passed to LLM that helps the agent to better plan and react given past interactions or data not immediately available.

There are three types of long term memory:

𝟭. Episodic - This type of memory contains past interactions and actions performed by the agent. After an action is taken, the application controlling the agent would store the action in some kind of persistent storage so that it can be retrieved later if needed. A good example would be using a vector Database to store semantic meaning of the interactions.

𝟮. Semantic - Any external information that is available to the agent and any knowledge the agent should have about itself. You can think of this as a context similar to one used in RAG applications. It can be internal knowledge only available to the agent or a grounding context to isolate part of the internet scale data for more accurate answers.

𝟯. Procedural - This is systemic information like the structure of the System Prompt, available tools, guardrails etc. It will usually be stored in Git, Prompt and Tool Registries.

𝟰. Occasionally, the agent application would pull information from long-term memory and store it locally if it is needed for the task at hand.

𝟱. All of the information pulled together from the long-term or stored in local memory is called short-term or working memory. Compiling all of it into a prompt will produce the prompt to be passed to the LLM and it will provide further actions to be taken by the system.

And that is it! The rest is all about how you architect the flow of your Agentic systems.

What do you think about memory in AI Agents?

#LLM #AI #MachineLearning

----------

Be sure to ♻️ repost if you found the article useful and follow me if you want to get a daily dose of useful AI related content in your feed!

[Petter Bae Brandtzaeg @PetterBB]

Vi går en raus tid i møte.

[Donald J. Trump @realDonaldTrump]

Happy Thanksgiving to all, including to the Radical Left Lunatics who have worked so hard to destroy our Country, but who have miserably failed, and will always fail, because their ideas and policies are so hopelessly bad that the great people of our Nation just gave a landslide victory to those who want to MAKE AMERICA GREAT AGAIN! Don’t worry, our Country will soon be respected, productive, fair, and strong, and you will be, more than ever before, proud to be an American!

[Jose @jarfbtc]

@Rothmus I hate LinkedIn.

[Petter Bae Brandtzaeg @PetterBB]

Welcome to the eighth CONVERSATIONS workshop on chatbots and human-centred AI. This will be an exciting event covering everything you need about human-AI interaction. Please Share