Local models, not only rented cloud
Custom GPU infrastructure makes testing, media processing and research easier to inspect.
TechJournal
Understand technology before it becomes everyday life.
TechJournal is the communication channel. AIJIM is the method. NamicGreen is the practice field. UTS is the research context. Together they create a view of AI, AI infrastructure, journalism and technology that does not only impress, but remains testable.
Current AIAI in journalismAI server & studioUTS researchNamicGreen / EXISTAIJIM
See the source. Test the claim. Keep the decision.
Episode 01
Why TechJournal? AI you can test.
The first episode is the viewer contract: why is it not enough when AI sounds convincing - and how does a strong claim become a testable analysis?
Deep Dive9-13 minEpisode in progress
Studio lab
Where technology becomes practical.
AI server, NAS, 10GbE, camera, lighting, voice and media production are not side content. They show what control, cost, privacy, latency and reproducibility mean in a real workflow.
Camera, lighting and explanation space
Front camera, studio atmosphere and B-roll make abstract AI questions visible.
From source to publication
Research, review, voice, editing and publishing are treated as one connected process.
Signature system
Every episode follows a clear trace.
TechJournal does not claim automated truth detection. Every analysis shows how a claim is framed - and where the decision remains human.
Claim
What exactly is being claimed - and for what purpose?
Evidence
Which source, measurement, observation or experience supports it?
Limit
What was not measured, not shown or not supported?
Decision
Which practical decision remains human and accountable?
Example claims
How TechJournal checks - without a truth machine.
No free input, no live verdict, no score. Just curated examples that show how editorial analysis works.
ClaimThe new model solves complex tasks autonomously.
EvidenceProduct demo, benchmark, own test or inspectable workflow?
LimitDemo setup, data, failure cases and human interventions often stay invisible.
DecisionFirst clarify which real task the model is robust enough for.
ClaimAI can speed up research and fact-checking significantly.
EvidenceWhich sources, documents and review steps remain connected?
LimitSummarization is not verification; a source is not automatically evidence.
DecisionPublication remains an editorial responsibility.
ClaimA local AI server makes AI work more controllable.
EvidenceCost, latency, privacy, data flow and reproducibility can be compared concretely.
LimitHardware does not solve method questions and creates its own setup limits.
DecisionLocal infrastructure pays off where control and workflow justify the effort.
Method
What TechJournal clarifies for you.
Every strong technology or AI claim is broken down editorially: what is being claimed, what evidence exists, what was not measured and where a human must decide?
Claim Evidence LimitChoose a topic in the orbit.
Who is behind it
Torsten Olivi Tiltack analyzes AI through research, studio and practice.
TechJournal explains. AIJIM provides the method. NamicGreen brings the practice field. UTS provides the research context. The studio lab makes servers, tools, hardware and media workflows concrete.
Research on Trustworthy AI, Data Science and AI-supported decision systems.
Springer conference chapter and arXiv preprint as publicly verifiable research core.
NamicGreen connects research, responsibility and real-world application fields.
AI server, camera, voice, tools, hardware and local AI workflows make technology visible.
PhD Candidate at UTS
Torsten Olivi Tiltack researches Trustworthy AI, Data Science and AI-supported decision systems at the University of Technology Sydney.
AIJIM as a method anchor
AIJIM is publicly listed as a Springer conference chapter and arXiv preprint. On TechJournal it is the method anchor for testable AI work, not the whole brand.
AI infrastructure, journalism and practice
Studio work, an AI server, tools, hardware, AI in journalism, NamicGreen and EXIST make abstract AI questions concrete: what can be built, tested and responsibly used?
Sources & Publications
Strong claims remain traceable.
TechJournal separates publicly verifiable research, locally checked experience, personal context and analysis. This keeps visible what is sourced, checked, interpreted or still open.
- AIJIM / FTC 2025publicly sourcedSpringer conference chapter on AIJIM for environmental journalism.
- AIJIM / arXivpublicly sourcedPublic preprint on AIJIM: Real-Time AI in Environmental Journalism.
- BIS 2026publicly sourcedEvidence-First Reference Model in Springer Proceedings / BIS programme.
- UTS profilepublicly sourcedTorsten Olivi Tiltack as Graduate Research Student / Higher Degree by Research.
- AIJIM Workspaceproject contextResearch and workspace logic for testable AI work.
- NamicGreen / EXISTpractice contextPractice and startup context; funding details stay out of public copy.
- LINAtechnical contextOnly framed as a technical voice system and studio topic.
TechJournal
When AI sounds convincing, the real question starts here.
What is being claimed? What evidence exists? What remains open? TechJournal makes technology understandable, relevant and testable.