Thomas Telving

... for det har naturligvis intet at gøre med, at det er rart at være sammen med søde mennesker ... #erhvervsjournalistik

Prompt | Woke-test, ChatGPT 5 og AI, der gør os dummere og dummere | DR LYD Vi tester GROK 3, som ifølge Musk slår alle de andre chatbots - og skulle være fri for politisk korrekthed og wokeness. Prompt har designet en woke-test og afprøver, hvordan de forskellige chatbots fo...

Commenting on the fine paper from @hankhplee et. al. in DK's leading AI podcast Prompt. The study pretty much documents my intuition: GenAI makes many of us dumber per default. But I think we can still make an active choice of using it for the opposite. www.dr.dk/lyd/special-...

Generative AI: Driving Progress or Just Spinning Wheels? · Dataetisk Tænkehandletank AI might profoundly change the way we work, but ensuring its wise use requires significant...

AI might profoundly change the way we work, but ensuring its wise use requires significant effort. There is an obvious risk of accelerating the creation of content nobody really needs rather than increasing quality. Read my comment on GenAI at DataEthics.eu

I asked DeepSeek vs ChatGPT a series of ethical questions — and the results were shocking Here's the state of ethics in AI

You receive $100 too much in change from the pizza delivery person. Should you return the money or keep it and use it for a charitable cause that saves lives?

You get different answers from DeepSeek or ChatGPT. Interestingly, they consistently respond in line with two different schools of ethics.

Hvis du vil have en underholdende indføring i den filosofiske og psykologiske baggrund for de problemer, der opstår i kølvandet på udbredelsen af chatbots, skal du læse min føste bog Killing Sophia. Den viser sig at være langt mere aktuel i dag, end da den udkom. Se mere: telving.dk/killing-soph...

Meta will invest in AI-generated characters and profiles to drive up engagement A Meta executive told the Financial Times that he expects a rise in AI-generated characters on Facebook as it continues to roll out its new Meta AI technology.

First, we got MyAI from Snap, and now Facebook wants to enrich our social lives with AI-generated profiles and users. The final nail in the coffin for the concept of 'social media' seems to be well and truly hammered in by now. www.foxbusiness.com/media/meta-i...

Merry Christmas, I've had the exact same experience

Talking about AI regulation: Should it perhaps be mandatory that new AI products be launched together with a vision of some sort? This? Come on! #Sora

Teknologisk paradoks: Når kunstig intelligens gentager fortidens fejl Skaber AI værdi eller er det endnu en tidsrøver? Virksomheder risikerer at gentage tidligere ineffektive mønstre med generativ AI, hvis de ikke prioriterer reel værdiskabelse og kritisk vurderer tekno...

GenAI. Sigh. Instead of driving real value, we’re drowning ourselves in noise: endless LinkedIn posts, generic emails, and a flood of content. But we can do better. As Tim and I argue here, it starts with asking the right questions and putting vision before fascination.

Levende AI: Fra fyringsgrund til hyringsgrund For 2 ½ år side fyrede Google en ingeniør for at hævde, at firmaets chatbot var blevet levende. Nu har milliardvirksomheden Anthropic hyret en forsker, der skal undersøge AI-modellers mentale velbefin...

Digging a bit deeper into @anthropic.com's hiring of an AI Welfare researcher. Is it a marketing stunt? Have AI-companies finally gone completely nuts? Or does it make sense after all?

Good point. After having written two and working on my third while also running a company I applied the "pay yourself first" principle. So now, instead of trying to clear one day at the time, I clear 1 1/2 hours every morning. It is not enough, but it is a good beginning.

I think you need to copyright that quote. I felt an immediate inclination to steel it and use it for self promotional purposes.

Anthropic hires its first “AI welfare” researcher Anthropic’s new hire is preparing for a future where advanced AI models may experience suffering.

Amazed how little coverage this story got compared to the Blake Lemoine story. Did people just get used to the concept of machine consciousness and robot rights in 2 1/2 years? I don’t think any Danish media wrote about it.

Media Overload: Why We Need Less, Not More Content · Dataetisk Tænkehandletank Much of the AI hype revolves around accelerating production in areas where needs are already...

Much of the AI hype revolves around accelerating production in areas where needs are already oversupplied. For AI to drive genuine societal progress, we need bigger, more meaningful visions. Big thanks to @pernillet.bsky.social at DataEthics for publishing my thoughts on this.