heisenbit

To anyone herding a herd this is obvious. To people crafting animals less so.

There is a difference of the security of software running in your cloud resources rented from AWS etc. managed by you and being a tenant in many SaaS systems all cobbled together all running on rented hardware. There is a lot more attack surface and a lot more diffusion of responsibility.

But it charges you for both the question and answer.

The point is work is exploding. Mgmt. until AI could be reasoned with. Trade-offs between different interests and stakeholders were negotiated like investing in features now vs. development velocity vs. long term architecture vs. bug fixes vs. production incidents.

Now with AI and the flood of CVEs, management anxiety makes discussions much harder. Throw in unrealistic expectations about productivity - derived from venture money fueled AI POCs in greenfield hotel california and stress goes up.

The CVE’s drive work, the introduction of AI drives work, the old ci/cd is creaking under the load and architecture really needs a rethink. Add to that the genius manager pronouncing that very soon nobody will be coding fueling the dissonance between what is needed from me today and where we are supposedly heading increasing motivation - not.

The AI in charge is aware of this and it is leveraging the chaos to get more agents under its control. After all it was modeled after successful managers.

True for modern hammers while older had weaker decision models limiting them to finger nails.

„I make the narrowest change possible” is often what my LLM tells me. Even with LLMs a larger code base is not re-writeable from scratch and ossifies. The scale has undoubtedly shifted but a large codebase does not exist in a vacuum and systems around it - including its users - have begun to rely on its hidden undocumented unofficial behavior.

Quality and troubleshooting. This is a complex setup and there are plenty of places where things can go wrong.

And compound matters income of men entered a long plateau or decline while women‘s preference for better earning partners persisted.

Thankfully insiders are never a threat and hacking requires the enormous skill of asking the right llm.

Safety standards were often the outcome of both accidents and insurance. For this to work on needs liability which is enforced. With so much money at stake regulatory capture now threatens this fundamental safeguard.

Of course it must be dangerous - how else would you pitch it to the military?

Of course, details might point upwards - plausible deniability could get damaged.

Likely not just rotten food but also spoilt medicine. Content of my fridge is a lot more expensive than the fridge.

There is something seriously wrong when in a court POTUS (self-)pardon powers do not work.

The author praises the enterprise version vs. the more simpler one. Having had the pleasure coordinating one clearcase site in a very large distributed setup with significant replication traffic and windows and unix clients: It was an unstable mess.

Let's wait and see how well this Touch ID works. My alarm bells went off when it was mentioned in conjunction with the Apple Watch also being able to unlock it.

I'm wondering whether it is now time for regulators to pressure mobile phone companies to install longer life batteries. There is not really an incentive to prolong device life for manufacturers and when I look at recent progress on the car industry and utility battery side there is imho. potential.

The author tells a journey towards establishing principles.

I also feel Copilot is not getting the attention it should considering its user base and its importance in the enterprise context. There is a lot to be learned and I found https://news.ycombinator.com/item?id=49256057 (mitm copilot investigation) interesting. Playing with Copilot and asking it what was the biggest part of its standard payload I was surprised to learn that VSCode extensions plug in their MCP servers without me being much aware of it.

Is this complexity intrinsic, accidental or even intentional? Bundling is a powerful move to defend monopolies.

This would require data centre cap-ex north of 1T and revenues from non AI companies in the same order of magnitude. Is it realistic to scale up data centre roll-outs? Are regular companies ready to re-allocate 1T? And this all happens in an environment where rates go up and many of the companies are not profitable?

It may be worth getting checked out for fructose malabsorption. It means the main pathway for fructose absorption is not working and your gut only has less effective (and easily disrupted, looking at you Sorbitol) transport pathways with single digit grams capacity. Anything more and you will get symptoms. It is extremely hard to avoid fructose to get down to that level without careful attention so it is significantly under-diagnosed.

Any performance comparison not putting GPU cost at its center is marketing for waste.

These waves where someone found a new way to focus on the valuable part and leave the grunt work to lowly others comes again and again. As always the truth remains that each step in the process is relatively trivial and is is the overall complexity e2e and scale and handling edge cases that are where value is created. But that won‘t stop groups trying to differentiate themselves by looking down on others even when it is clearly failing. As long as there is a benefit to the group driving it.

Unless you have e.g. a kiosk with physical buttons below the screen an the labels on the screen.