Savlar4 sources healthy · latest episode 15 Oct 2026
Speaker

Alexander Mattick

Guest on Machine Learning Street Talk (MLST). 568 claims across 1 episode, Sept 2026.

568claims, each timestamped
1 of 12signals on the brief
0verified changes of mind
31open predictions

Positions over time

Every pair of statements that was compared. Only a verified change counts as a change of mind; rewording, added nuance and different questions do not.

No two statements by this speaker have been compared yet.

Predictions

As stated. A window appears only where the speaker gave one.

ExpectsBySaidStatus
A repulsive effect can eventually emerge if all points move toward the density's mode, while the fixed-density approach retains flexibility. no date given21 Sept 2026pending
Agentic systems could benefit from using a Bayesian approach in which they gather data. no date given21 Sept 2026pending
The speaker highly doubts that, five years from the episode date, the startup's work will qualify as an energy-based model under the current understanding, even if it is labeled one. 21 Sept 2031 – 21 Sept 203121 Sept 2026pending
General models will probably exhibit a similar pattern in which coarse structure emerges before finer details. no date given21 Sept 2026pending
Large-scale structure will emerge before smaller details when successive latent changes are constrained to be small. no date given21 Sept 2026pending
A diffusion model constrained to produce small details before large-scale structure might require thousands of times more parameters to fit the data. no date given21 Sept 2026pending
The speaker expects most diffusion models to produce large-scale structure before details, while acknowledging that a model could be designed to produce small details first. no date given21 Sept 2026pending
Alexander Mattick allows that one model architecture might scale more efficiently than another. no date given21 Sept 2026pending
Higher-information features will probably appear first in diffusion models because they are easier gains, though Alexander Mattick does not consider this necessary. no date given21 Sept 2026pending
The speaker expects function-space analysis to be more useful than parameterized analysis in the long term because it can apply across architectures, while parameterized analysis explains specific architectures. no date given21 Sept 2026pending
Function-space analysis would remain applicable if architectures shifted from transformers to MLP mixers. no date given21 Sept 2026pending
The speaker believes that theory will never explain literally everything. no date given21 Sept 2026pending
The machine-learning field will remain unprepared to produce a unified theory comparable to physics until the pace of progress slows. no date given21 Sept 2026pending
The speaker says Sutton's view implies that becoming very good at reward maximization would make research into other complicated approaches obsolete. no date given21 Sept 2026pending
Removing the negative reward for killing an enemy hero would lead the agent to repeatedly kill that hero in the middle of the map rather than make progress. no date given21 Sept 2026pending
A robotic system without strong guarantees might crush someone. In a cluster of 3 →no date given21 Sept 2026pending
If models generalize, planning with a sufficiently large world model could enable the task under discussion, provided compute and latency are not limiting and the relevant task-specific data is available. no date given21 Sept 2026pending
If models generalize, planning with a sufficiently large world model could enable robots to perform tasks, provided compute and latency are not limiting and relevant task data is available. no date given21 Sept 2026pending
Nonlinear model predictive control (NMPC) will ultimately provide a solution; Alexander Mattick considers this possible. no date given21 Sept 2026pending
Alexander Mattick suspects that world models will face limitations analogous to those of optimization-based control, though they may solve enough easy tasks to suffice for many applications. no date given21 Sept 2026pending
The speaker expects world models to handle many relatively easy tasks, and thinks that capability may already suffice for many tasks. no date given21 Sept 2026pending
The speaker expects world models to face severe limitations due to real-world conditions and the diversity and complexity of tasks, while not necessarily being fundamentally flawed as an idea. no date given21 Sept 2026pending
Alexander Mattick suspects that world-model approaches will encounter hard limits due to real-world conditions and the diversity and complexity of tasks, though not necessarily due to the underlying idea. no date given21 Sept 2026pending
Alexander Mattick thinks an NMPC controller could plan a simple manipulation task, such as using the other hand to move an object when stuck, if its model is sufficiently accurate. no date given21 Sept 2026pending
An NMPC controller could plan to use the other hand to move a stuck ball, provided its model is accurate enough. no date given21 Sept 2026pending
Alexander Mattick does not expect home robots to be deployed initially or in the short-to-medium term, and is uncertain about whether they will be deployed in the long term. no date given21 Sept 2026pending
The speaker does not expect home robots to be deployed initially or in the short to medium term, and is uncertain whether they will be deployed in the long term. no date given21 Sept 2026pending
A company such as BMW would reject a robot that performs correctly only 90 percent of the time and fails or does something else 10 percent of the time. no date given21 Sept 2026pending
Alexander Mattick argues that an industrial customer such as BMW would not pay for a robot that performs correctly only 90 percent of the time. no date given21 Sept 2026pending
An industrial customer might prefer investing in C/C++ code optimization if it yields greater effective speed than buying an unreliable robot. no date given21 Sept 2026pending
The company might prefer to spend the money fine-tuning its C and C++ code, which could provide more effective speed. no date given21 Sept 2026pending

Appearances

Claims per episode