PunditScore
Pundits › Sebastian Raschka, PhD
SP

Sebastian Raschka, PhD

Compare →
LLM Research Engineer, Author, Machine Learning Expert
Operational & industryArtificial IntelligenceMachine Learning

Sebastian Raschka, PhD, is an LLM Research Engineer with over a decade of experience in artificial intelligence, specializing in the development of high-performance AI systems and large language models. He has previously served as an assistant professor in the Department of Statistics at the University of Wisconsin-Madison and has authored several bestselling books on machine learning. Currently, he works at Lightning AI, focusing on implementing and training LLMs.

Score over time

The PunditScore after each graded call, on the panel's score bands.
Well above cohortAbove cohortAt cohort averageBelow cohortWell below cohort30405060708090#1#3#5#7#9#11#13Panel average per prediction — 56.3RIGHT2022-11-04 — Sign in to read this call.2023-10-23 — Sign in to read this call.2025 — Sign in to read this call.2025-01-27 — Sign in to read this call.2025-01-27 — Sign in to read this call.2025-01-27 — Distilling down larger models will become the default for many researchers and practitioners interested in 'small' LLM model training.2025-01-27 — Sign in to read this call.2025-01-27 — Sign in to read this call.2025-12-09 — Training a Qwen3 model with reinforcement learning and verifiable rewards may be a good candidate for 2025.2025-12-30 — The open-weight community will slowly but steadily adopt LLMs with local tool use and increasingly agentic capabilities.2025-12-30 — RLVR will more widely expand into other domains beyond math and coding (for example, chemistry, biology, and others).2025-12-30 — Classical RAG will slowly fade as a default solution for document queries. Instead of using retrieval on every document-related query, developers will rely more on better long-context handling.PARTLY2022-10-15 — Sign in to read this call.2022-11-04 — Sign in to read this call.OPEN2022-10-15 — Sign in to read this call.2025 — Sign in to read this call.2025 — Sign in to read this call.2025-01-27 — Sign in to read this call.2025-04-07 — We will likely also see Llama 4.1 (and perhaps Llama 4.2).2025-12-30 — A lot of LLM benchmark and performance progress will come from improved tooling and inference-time scaling rather than from training or the core model itself.2025-12-30 — AI will transition from a 'logical deductive engine' to a 'negotiator of meaning.'2025-12-30 — The ecosystem will stabilize in 2026-2027.2026-02-01 — Big acquisitions in the AI sector will occur in 2026.
Hover a point to read the call. Click to pin it.
AXIS

Still open · 9

THE FRONT EDGE

Not resolved, so they don’t move the score.

?
Big acquisitions in the AI sector will occur in 2026.
Due Dec 31, 2026Said February 1, 2026AIAcquisitions
No market prices this claim
?
AI will transition from a 'logical deductive engine' to a 'negotiator of meaning.'
Due Dec 31, 2026Said December 30, 2025AILLMs
No market prices this claim
?
Sign in to read this call.
Due Dec 31, 2026Said 2025AILLMs
No market prices this claim
?
Sign in to read this call.
Due Dec 31, 2026Said 2025AILLMs
No market prices this claim
?
Sign in to read this call.
Due Dec 31, 2026Said October 15, 2022LLMsAdvancements
No market prices this claim
?
The ecosystem will stabilize in 2026-2027.
Due Dec 31, 2027Said December 30, 2025AIEcosystem
No market prices this claim
?
A lot of LLM benchmark and performance progress will come from improved tooling and inference-time scaling rather than from training or the core model itself.
Said December 30, 2025AILLMs
No market prices this claim
?
We will likely also see Llama 4.1 (and perhaps Llama 4.2).
Said April 7, 2025AIMachine Learning
No market prices this claim
?
Sign in to read this call.
Said January 27, 2025LLMopen-source
No market prices this claim