Hotblack
@invalid-handle.com
Will talk science, tech, economics, politics
Mainly post stupid puns and captions for pictures
Cares about inequalities and exclusion, particularly around health
Enjoys cricket
Mainly post stupid puns and captions for pictures
Cares about inequalities and exclusion, particularly around health
Enjoys cricket
Looking forward to the first tabloid to leave purely so it can trash all its rivals with alliterative headlines using words begging with p
January 13, 2026 at 5:02 PM
Looking forward to the first tabloid to leave purely so it can trash all its rivals with alliterative headlines using words begging with p
Agreed. For the orchestration
I’m thinking more about training models. Are remaining shortcomings in LLM based agents because they have not learnt from experience but from ingesting data?If they could learn, from scratch, how to program… perhaps they would be more capable, creative and verifiable?
I’m thinking more about training models. Are remaining shortcomings in LLM based agents because they have not learnt from experience but from ingesting data?If they could learn, from scratch, how to program… perhaps they would be more capable, creative and verifiable?
January 13, 2026 at 4:56 PM
Agreed. For the orchestration
I’m thinking more about training models. Are remaining shortcomings in LLM based agents because they have not learnt from experience but from ingesting data?If they could learn, from scratch, how to program… perhaps they would be more capable, creative and verifiable?
I’m thinking more about training models. Are remaining shortcomings in LLM based agents because they have not learnt from experience but from ingesting data?If they could learn, from scratch, how to program… perhaps they would be more capable, creative and verifiable?
I’m really interested in what will happen when the coding agents are less LLMs and more self taught through reinforcement learning. Removing the human and learning to program from scratch, like AlphaGo or AlphaProof did in their fields
Does Doll have views on this?
Does Doll have views on this?
January 13, 2026 at 4:47 PM
I’m really interested in what will happen when the coding agents are less LLMs and more self taught through reinforcement learning. Removing the human and learning to program from scratch, like AlphaGo or AlphaProof did in their fields
Does Doll have views on this?
Does Doll have views on this?
Thanks! 🙂 May be a bit more intermittent than previously but we will see
January 13, 2026 at 4:33 PM
Thanks! 🙂 May be a bit more intermittent than previously but we will see
A more serious and informative take on all this from @darioperkins.bsky.social and co
Powell vs Trump - It’s WAR — GlobalData TS Lombard: Perkins Vs Beamish
Andrew Slazenger (reach out for a free trial, institutional investors only: [email protected]) hosts Dario Perkins and Freya Beamish as they answer:1. What are our new year resolutions?2....
overcast.fm
January 13, 2026 at 3:44 PM
A more serious and informative take on all this from @darioperkins.bsky.social and co
All things are transient…. even Gordon Way
January 13, 2026 at 3:07 PM
All things are transient…. even Gordon Way
I’m looking forward to the publications of Quotations from Chairman Trump
The little red, white and blue book?
The little red, white and blue book?
January 13, 2026 at 2:00 PM
I’m looking forward to the publications of Quotations from Chairman Trump
The little red, white and blue book?
The little red, white and blue book?
I bet he’s got a gravelly purr
January 12, 2026 at 8:29 PM
I bet he’s got a gravelly purr
Thanks Ben, appreciated 😊
December 11, 2025 at 4:23 PM
Thanks Ben, appreciated 😊