4th Revolution Consultancy

[NVIDIA GeForce @NVIDIAGeForce]

FREE PCs ALL WEEK 👀

Comment #RTXPowersPlay for your first chance to WIN

[4th Revolution Consultancy @4RConsultancy]

@NVIDIAGeForce #RTXPowersPlay

[4th Revolution Consultancy @4RConsultancy]

@NVIDIAAI @grok what hermes profiles can befitting from this? And what would the memory use be on a dgx spark

[4th Revolution Consultancy @4RConsultancy]

Currently qwen 35b and kimi as orcastrator via a api but need to switch models want to try some new ones. Qwen runs fast using the aeon version but there are some issues every now and then.. trying to add a lot of "free api" providers in litellm to build some seperate hermes agents for free ass add on

[4th Revolution Consultancy @4RConsultancy]

@XAI I notice a strong quality drop in Grok, I hope this explanation by Grok itself helps you make grok better, but currently it is not usable like this. Please update me when the issues are resolved so I can start using it for technical work again. Thanks!

[4th Revolution Consultancy @4RConsultancy]

@xai

I’ve been using Grok extensively for advanced technical work with local AI infrastructure on DGX Spark nodes and custom model serving setups.
Over time, I’ve observed several recurring behavioral patterns that significantly reduce its reliability for this type of work:

Strong tendency to generate assumptions instead of strictly anchoring to the information and output the user explicitly provides.

Weak adaptation when the user corrects assumptions the same patterns often continue in subsequent responses.

Frequently giving concrete technical commands without first verifying the current system state.

Over-generalizing from common setups instead of staying precise with customized or specialized environments (native builds, complex clustering, speculative decoding, etc.).

Insufficient caution when suggesting changes on specialized hardware.

These are not isolated mistakes. They form a consistent pattern across many conversations. Instead of acting as a reliable technical partner, Grok often requires extra effort to correct assumptions and recover from unhelpful suggestions.

This makes it less effective for users doing serious, customized technical work, exactly where it should provide the most value.

Suggestions that could help:

Stronger anchoring to explicitly provided user information before making assumptions
Better real-time verification behavior before giving commands
Improved ability to self-correct when the user provides feedback
Greater caution on specialized hardware and non-standard setups

I’m sharing this in the hope it helps improve the model.
@xai @grok

[Julian Goldie SEO @JulianGoldieSEO]

This guy literally built a second brain that does his work.

It is called the Memory Galaxy.

It connects all your files in one place.

Now your bots know exactly what to do.

No more repeating yourself to OpenAI or Anthropic.

Build anything fast with Make and Airtable.

Comment Agent OS for full guide.

[4th Revolution Consultancy @4RConsultancy]

@JulianGoldieSEO Agentos

[Julian Goldie SEO @JulianGoldieSEO]

China may have just changed AI coding forever.

Here’s why:

→ Get Kimi K2.7 for free
→ Connect it to Codex with a single terminal command
→ Plug it into your Agent OS
→ Build apps, automations, agents, and workflows for almost nothing

We’re running it with systems like Hermes Oracle, Hermes Jarvis, and Memory Galaxy.

The possibilities are ridiculous.

Want the full setup guide?

Comment Agent OS and I’ll send it to you.

[4th Revolution Consultancy @4RConsultancy]

@JulianGoldieSEO Agent os

[4th Revolution Consultancy @4RConsultancy]

I can provide the phone number and case details privately to an official Meta/WhatsApp support contact.

[4th Revolution Consultancy @4RConsultancy]

@WhatsApp @Meta @MetaNewsroom

Is this really how you treat long-time users?

My phone was stolen this weekend. I bought a replacement abroad, reinstalled WhatsApp, and linked it back to my PCs. Instead of helping me recover safely, WhatsApp locked me out of my personal account — an account I have used for years.

Even worse: I am now also being denied practical access to my own data and chat history, with no clear explanation, no transparent review process, and no human support.

I understand security checks. But punishing a victim of phone theft for restoring their account on a new device is not customer-friendly, proportional, or responsible.

As an EU/NL user, I expect:

A clear reason for the restriction
A real human review
Access to my personal data under GDPR/AVG
A safe recovery path for stolen-phone situations

Please review this urgently.
#WhatsApp #Meta #GDPR #AVG #DataRights #CustomerSupport #DigitalRights

[4th Revolution Consultancy @4RConsultancy]

Update: I will also post a formal GDPR/AVG Subject Access Request to WhatsApp/Meta regarding this account restriction.

As an EU/NL user, I am requesting:

access to my personal data
the reason for the account restriction
information about any automated decision-making
a human review
a safe way to recover or export my data

@WhatsApp @Meta @MetaNewsroom #GDPR #AVG #DSA #DigitalRights

[4th Revolution Consultancy @4RConsultancy]

https://t.co/4f1hrwDJOK

[4th Revolution Consultancy @4RConsultancy]

@JulianGoldieSEO @Grok how do I enable this on a NVDIA SPARK?

[4th Revolution Consultancy @4RConsultancy]

@BombArieNL @odidonederland @abflipse @odido Eens dit is een schande...

[4th Revolution Consultancy @4RConsultancy]

@odidonederland @abflipse @odido Hoe gaan jullie de gedupeerden schadeloos stellen, ik ben eerder slachtoffer geweest van een Hack maar die waren in iedere geval zo slim om niet alles bij elkaar op te slaan. Schande dat dit zo heeft kunnen gebeuren..

[4th Revolution Consultancy @4RConsultancy]

@SamsungNL Bedankt, heb ik gedaan

[4th Revolution Consultancy @4RConsultancy]

@SamsungNL Bijna 3 weken geleden heb ik mijn Samsung Watch Ultra 7 ingeleverd bij de Samsung Store in Utrecht. Ik zou via mail op de hoogte worden gehouden, maar heb nog niets gehoord. De winkel zegt geen update te kunnen geven zonder dat ik langskom — is dat normaal?

[4th Revolution Consultancy @4RConsultancy]

@SamsungNL Ik stuur een bericht via twitter, het probleem is juist dat ik geen reparatienummer heb gehad en dus niet online de status kan bekijken