@wordshaper @bsletten Any #AI that thought they were going to monetize and capitalize are fucked. When Google.gemini started censoring output about #White #Conservative #USPoliticians and trying to charge, I simply started using #NonUSAI #DeepSeek #Cohere #Mistral #Claude
"Anthropic is very likely losing money on every single Claude Code customer, and based on my analysis, appears to be losing hundreds or even thousands of dollars per customer.
There is a gaping wound in the side of Anthropic, and it threatens financial doom for the company.
Some caveats before we continue:
- CCusage is not direct information from Anthropic, and thus there may be things we don’t know about how it charges customers, or any means of efficiency it may have.
- Despite the amount of evidence I’ve found, we do not have a representative sample of exact pricing. This evidence comes from people who use Claude Code, are measuring their usage, and elected to post their CCusage dashboards online — which likely represents a small sample of the total user base.
- Nevertheless, the amount of cases I’ve found online of egregious, unrelentingly unprofitable burn are deeply concerning, and it’s hard to imagine that these examples are outliers.
- We do not know if the current, unrestricted version of Claude Code will last.
The reason I’m leading with these caveats is because the numbers I’ve found about the sheer amount of money Claude Code’s users are burning are absolutely shocking.
In the event that they are representative of the greater picture of Anthropic’s customer base, this company is wilfully burning 200% to 3000% of each Pro or Max customer that interacts with Claude Code, and in each price point’s case I have found repeated evidence that customers are allowed to burn their entire monthly payment in compute within, at best, eight days, with some cases involving customers on a $200-a-month subscription burning as much as $10,000 worth of compute."
We do not provide evidence that:
- #AI systems do not currently speed up many or most #softwareDevelopers
- AI systems do not speed up individuals or groups in domains other than #softwareDevelopment
- AI systems in the near future will not speed up #developers
- There are not ways of using existing AI systems more effectively
> Measuring the Impact of Early-2025 AI on Experienced #OpenSource #Developer #Productivity
https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
"We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers. The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't."
"we find that when developers use AI tools, they take 19% longer than without - AI makes them slower." [1]
- says a study with a randomized and controlled trial, which includes a chart of hilariously overly optimistic forecasts against woefully observed results
On a very simple image of three arrows at the top left, top centre, and top right of the screen it takes 61 seconds using the traditional raster engraving method. Using the new image segmentation method it takes just 18 seconds. A 70% improvement.
I’m just fixing the final bugs then will get it uploaded to GitHub and make a video.
Trying to add image segmentation to K40 Whisperer, this will massively speed up some engraving jobs where the thing to engrave is small but scattered throughout the page.
It feels quite scary using Claude Code and Vibe Coding to control something physical such as the X/Yposition of a laser head.
It’s working, I just have a few bugs to resolve and then I will post some videos.
wenn man, wie bei #grok, sieht wie #KI manipuliert werden kann oder, wie bei #claude, bestimmte verhaltensweisen nicht nachvollziehen kann, dann ist es eine verdammt dumme idee solche technologie für militärische zwecke einzusetzen
https://breakingdefense.com/2025/04/nato-picks-palantirs-maven-ai-for-military-planning-amid-trans-atlantic-tension/
All the Frontier Model Companies are Now offering Significant Educational Services https://stefanbauschard.substack.com/p/all-the-frontier-model-companies #AI #education #Claude #OpenAI #Google
I really liked this post and they nailed several points I have struggled to find the time to write down.
I have been telling people this for months:
> Cursor, Windsurf, and Copilot all started with VS Code and diverged in various ways, but they were all bolting AI onto an editor that was designed pre-AI. Fundamentally, VS Code does not look very different from a JetBrains IDE from 20 years ago.
AI failed while in charge of a real business, underperforming human managers. The experiment shows, however, that AI learns fast and adapts well to new situations.
https://www.news-cafe.eu/?go=news&n=13706
#ai #artificialintelligence #business #economy #anthropic #claude @anthropic @andonlabs
It's not a coincidence that during a time when governments want you to think less the market is flooded with tools that do your thinking for you.
Getting #Claude Code to do Exactly What You Ask : #Hooks
Claude #Code’s “hooks” feature provides developers with precise control over coding tasks, allowing automation, rule enforcement, and error minimization.
Hooks act as predefined rules or guardrails, making sure code adheres to specific standards like naming conventions, architectural guidelines, and testing protocols.
https://www.geeky-gadgets.com/flawless-ai-code-generation-with-claude-code-hooks/
enjoying learning scottish flute music with the ABC app I built with the help of #Claude :)
The full #Anthropic post on their experience with #Claude running a vending machine is chock full of amusingly crazy behaviors.
However, this paragraph near the end exemplifies how we are stuck in an #AIBubble, because once again someone says "we experimented with having AI do a task, and it completely failed, but we will continue to believe that AI will be great at this task because surely someone will figure out how to get AI to do this task."
Gus Posey 
