My daily column. Where machines learn to speak, policy fails to keep up, and everything else gets the treatment it deserves.
status: In Progress
Status Indicator
The status indicator reflects the current state of the work:
- Abandoned: Work that has been discontinued
- Notes: Initial collections of thoughts and references
- Draft: Early structured version with a central thesis
- In Progress: Well-developed work actively being refined
- Finished: Completed work with no planned major changes
This helps readers understand the maturity and completeness of the content.
·
certainty: likely
Confidence Rating
The confidence tag expresses how well-supported the content is, or how likely its overall ideas are right. This uses a scale from "impossible" to "certain", based on the Kesselman List of Estimative Words:
1. "certain"
2. "highly likely"
3. "likely"
4. "possible"
5. "unlikely"
6. "highly unlikely"
7. "remote"
8. "impossible"
Even ideas that seem unlikely may be worth exploring if their potential impact is significant enough.
·
importance: 8/10
Importance Rating
The importance rating distinguishes between trivial topics and those which might change your life. Using a scale from 0-10, content is ranked based on its potential impact on:
- the reader
- the intended audience
- the world at large
For example, topics about fundamental research or transformative technologies would rank 9-10, while personal reflections or minor experiments might rank 0-1.
Topics1/3
May 2026
posted on 05.11.2026
Every week another AI startup raises a Series B on vibes and a demo that crashes if you sneeze near it. The valuations are untethered from revenue in a way that would make 1999 blush. I'm not saying the technology isn't real — it is — but the gap between what these models can do and what these companies are promising is a canyon you could lose a pension fund in.
The tell is always the same: when the press release says "enterprise-ready" but the API docs say "beta."
aieconomicsventure-capital
No reactions yet
June 2025
posted on 06.01.2025
The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem ComplexityBrydon Eastman, Chen Huang, Skyler Seto, Hadi Pouransari, Mehrdad Farajtabar, Raviteja Vemulapalli, Fartash Faghri, Oncel Tuzel, Barry-John Theobald, Josh SusskindJul 18, 2025
Anthropic launched Haiku, Sonnet, and Opus under the Claude 3 family, offering powerful capabilities across tasks.
Background
This section provides context and background information about Claude 3 Family.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: March 3, 2024
Related
Links to related articles, papers, or announcements.
Tags
Anthropic
Claude 3
Sonnet
Opus
Haiku
AnthropicClaudeLanguage Models
No reactions yet
February 2024
posted on 02.18.2024
Gemini 1.5 Pro released this week with the promise.
"The model delivers dramatically enhanced performance, with a breakthrough in long-context understanding across modalities."
It was designed to be a mid-size multi modal model that matches the performance of 1.0 Ultra (their largest model) while simultaneously managing to use less compute than the prized heifer. 1.5 uses a transformer, and mixture of experts architecture. MoE allows the model to be split into smaller "expert" narrow llms rather than the traditional monolith neural net. Meaning for any given input, only relevant expert pathways active, leading to more effective training and inference.
The defining feature of 1.5 Pro is still it's context window however.
Metric
Value
Standard context window
128,000 tokens
Max context window (preview)
1 million tokens
Tested in research up to
10 million tokens
A context window of 1 million tokens is equivalent to 1 hour of video, 11 hours of audio, >30K lines of code, >700K words.
The defining feature of 1.5 Pro is its context window:
Metric
Value
Standard context window
128,000 tokens
Max context window (preview)
1 million tokens
Tested in research up to
10 million tokens
What 1 million tokens can hold:
1 hour of video
11 hours of audio
30,000+ lines of code
700,000+ words
GoogleGeminiLanguage Models
No reactions yet
February 2024
posted on 02.15.2024
Gemini 1.5
Overview
Gemini 1.5 Pro featured industry-leading long-context capabilities, enabling advanced reasoning over vast documents.
Background
This section provides context and background information about Gemini 1.5.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: February 14, 2024
Related
Links to related articles, papers, or announcements.
Tags
Google
DeepMind
Gemini 1.5 Pro
Long Context
GoogleGeminiLanguage Models
No reactions yet
December 2023
posted on 12.06.2023
Gemini 1
Overview
The first Gemini model from Google DeepMind, merging strengths from AlphaCode, Pathways, and large-scale training.
Background
This section provides context and background information about Gemini 1.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: December 5, 2023
Related
Links to related articles, papers, or announcements.
Tags
Google
DeepMind
Gemini 1
GoogleGeminiLanguage Models
No reactions yet
November 2023
posted on 11.06.2023
GPT-4 Turbo
Overview
An optimized and cost-efficient variant of GPT-4 powering ChatGPT with custom GPTs, tools, and longer context.
Background
This section provides context and background information about GPT-4 Turbo.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: November 5, 2023
Related
Links to related articles, papers, or announcements.
Tags
OpenAI
GPT-4 Turbo
ChatGPT
OpenaiGptLanguage Models
No reactions yet
July 2023
posted on 07.11.2023
Claude 2
Overview
An improved Claude model with stronger reasoning, fewer hallucinations, and increased openness for public use.
Background
This section provides context and background information about Claude 2.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: July 10, 2023
Related
Links to related articles, papers, or announcements.
Tags
Anthropic
Claude 2
LLM
AnthropicClaudeLanguage Models
No reactions yet
March 2023
posted on 03.14.2023
GPT-4
Overview
A multimodal leap forward for OpenAI, capable of reasoning over images and text with more nuanced capabilities.
Background
This section provides context and background information about GPT-4.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: March 13, 2023
Related
Links to related articles, papers, or announcements.
Tags
OpenAI
GPT-4
Multimodal
OpenaiGptLanguage Models
No reactions yet
June 2020
posted on 06.11.2020
GPT-3
Overview
The 175B parameter model that revolutionized natural language interfaces and powered the first wave of AI API tools.
Background
This section provides context and background information about GPT-3.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: June 10, 2020
Related
Links to related articles, papers, or announcements.
Tags
OpenAI
GPT-3
API
OpenaiGptLanguage Models
No reactions yet
February 2019
posted on 02.14.2019
GPT-2
Overview
GPT-2 demonstrated surprisingly coherent text generation, sparking debate over AI safety and open-sourcing.
Background
This section provides context and background information about GPT-2.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: February 13, 2019
Related
Links to related articles, papers, or announcements.
Tags
OpenAI
GPT-2
Text Generation
OpenaiGptLanguage Models
No reactions yet
June 2018
posted on 06.11.2018
GPT-1
Overview
OpenAI's first generative pre-trained transformer model, laying the foundation for large language models.
Background
This section provides context and background information about GPT-1.
Key Details
Important details and specifics about this development.
Significance
Why this development is important in the broader context of AI and technology.
Timeline
Date: June 10, 2018
Related
Links to related articles, papers, or announcements.