I used their AI chat last night and I was impressed with the product implementation. I was able to use it to make quick sense of data and even generate prompts to solve problems locally.
IDK what the prior AI behaviors have been, but for me, what they have now is an ~idealized version of AI product.
It is basically what I'd do if they didn't offer it but not as well: export data, import their docs in md., import some industry best practices in analytics into context etc.
Except its all right there. Not sure of the economics of it, as it was on a free account and some reasonably good model was powering the discussion. But it was about as engaged as I've been with an analytics tool.
I think it is a mix of the sibling replies here. I'd add that the company has seemed to find ways to ~do more with less.
I have never liked the various nerfs Anthropic has used to balance GPU (slowing down responses, quota variance, model optimizations etc) and it definitely has burned a lot of good-will.
But it has seemed that being able to look beyond the short term pitchforks has worked quite well.
I also think this undersells the real value of the bot, which is to handle tasks via voice that an average human either would not or could not do.
In the video example with the grannies, the knitter is essentially wanting a PA. Regular folks don't have PAs. Even when that became a thing in the aughts they were all outsourced.
When I've used voice chat, it has often turns into rabbit holes on very niche topics. For example, I had one start about the 1996 performance of Rage Against the Machine in Portland, Oregon that was supposed to feature Wu Tang Clan. (already outside most human's knowledge) that dove into details of the club scene in Los Angeles at the time of RATM's signing to Epic Records.
Was anyone else here at that '96 show in Portland? It seems like it might be challenging to find a person on the internet able to engage on the topic.
The person may exist, but not during my fleeting interest in the subject while walking to the park.
How are y'all carrying context history from one agent to the other?
I also flip between the models due to quota, TUI enhancements, model updates and service availability.
To handle this, I built a thing that normalizes your transcripts between Claude Code and Codex into a shared DB, then a CLI and skill.
It has made it so it doesn't matter what I built where (or when) I just refer to the work and drop in a /total-recall (or $total-recall on codex) and the agent brings it into the current convo.
I realize there are a lot of ~memory tools out there, but I think particular my approach and product behavior is unique.
I’ve written a skill for codex and Claude code that designates an orchestrator on the primary worktree and is agnostic about what type of AI workers are on the N supporting worktrees.
The orchestrator knows which AI client is running in any given worktree, so it would be fairly easy to designate which AI should receive what kind of tasks.
You run either Claude or Codex in tabs for each work tree.
I do have some AI TUI specific instructions, for instance codex is primitive at monitoring compared to CC. So, there are additional notes for Codex workers on how to properly monitor for new "mail."
You work with the orchestrator on the primary worktree and allow it to delegates tasks to the workers and answer their smaller questions.
It surfaces results and assisting them with context clearing when needed.
The orchestrator and workers communicate using a simple shared file system under tmp/* and together they can handle a big and varied workload.
I use iterm2, so I’ve also added iterm2 specific python that allows the orchestrator to “kick” a worker or perform tasks otherwise veto'd by the TUIs (ie /clear) by modifying the input and submitting it.
I would attribute Disney's use of scarcity as a primary means to drive film and TV box office and streaming dollars in the Star Wars franchise.
This is already under threat due to the Star Wars AI videos being released on Youtube, seemingly without constraint as of yet.
The videos are not Hollywood quality [0], however they circumvent rules Disney can't easily break like using the likeness of any actor at any age in any circumstance.
These fan made videos get lots of views. Even if they were all removed from YouTube, this will be a difficult thing to stop.
I believe a generally accepted "good" or even "great" unofficial, Star Wars film built without sets or actors using AI is inevitable. And that this will be true for any popular franchise.
The natural corollary to this arc is into games, where using AI to code most or all of a AAA-competitive title would be considered inevitable.
I suspect Disney and Sony have at least someone pointing at this outcome.
[0] I suppose idealized Hollywood quality. They are better than some films.
It would be good if Anthropic provided some kind of feedback or even toggle to auto-route requests for models being used at thinking levels that would be a better value using a different model.
Sort of like, getting an automatic upgrade at a car rental or hotel if there is availability.
In that, it seems sonnet 5 on high costs more than opus 4.8 at a lower pass rate. Am I reading this correctly?
Edit: It looks like the key value proposition of the updated model is that it is much better than Sonnet 4.6.
Wheras, Sonnet 5 delivers great value (by browsercomp benchmarks and compared to opus) when running in low and medium.
So: Sonnet 4.6 should ~never have been run for low, medium or high when Opus 4.8 has been available. Whoops, I think I have some skills that delegate easy stuff to Sonnet.
---
I remember Anthropic pivoting everyone's default model to Opus but had not seen it put so starkly before.
I am a bit confused on the subscription `/usage` screen. It splits out sonnet usage, and I'd presumed that would have contributed to a lower use of subscription Quota.
But if this is correct, Sonnet usage was basically like smoking unfiltered cigarettes.
Interesting approach. I am curious how this information stacks up over time and how efficient it is at incorporating decision knowledge into active context.
I have taken a different approach: allow team members to sync all of their Claude Code and Codex transcripts on a project and give them a skill that lets them ask their AI why decisions were made.
The skill I've built, /total-recall is backed by a Swift-based CLI that provides efficient query tooling that coding agents can use however they see fit to arrive at the answer.
The corpus of data contextify queries is a SQL database managed by macOS and Linux clients. These clients ingest the jsonl files in realtime and optionally can sync transcript data through either a hosted or self-hosted server.
This allows any team member to simply invoke the skill: "Why did we switch over to allauth from aws cognito? /total-recall."
My experience is that Claude Code and Codex don't just land "near" a decision, but can assemble it from what is sometimes a winding pathway of research, benchmarking and experimentation.
Rather than codify requirements into a separate spec, Contextify lets agents pair the state of the code with the conversational record.
I have just released the free personal and source available self-hosted version of Contextify, I'd be glad for feedback.
When I designed our take on it, I was solving a problem I experienced on Craigslist. I had not seen this prior art.
I built a simple refresh for a new email address interface that people really loved to mash, and it is nearly identical to the Use Different Address link behavior on Hide My Email.
To get to my original point, if Craigslist was aware of all of these examples, they did not seem to serve as impetus to provide it, despite it being in the best interest of their users.
I would highlight again that the system described by the Rally patent, if realizable in the example services means these groups also left potentially valuable IP on the table.
As the lawsuit over Hide My Email, afaik, is serious stuff.
I appreciate folks sharing links to prior art. I have more to say, that might explain my initial comment a bit more, but have to wait on that.
Recent public projects include:
- Contextify: https://contextify.sh (A tool to assist CLI-based agentic programming workflows)
– FileKitty: https://github.com/banagale/FileKitty (a prompt engineering utility)
– Chief of Staff: https://chiefofstaffhq.com (a pre-generative AI text-to-speech SaaS)
I also very occasionally write at: https://banagale.com
Feel free to say hello, rob @ the domain above.