Muhammad Abdullahi dattijo

AI developer | Cybersecurity researcher | Technical writer | Exploring the systems shaping AI, security, and digital policy.

[Towards Data Science @TDataScience]

What, exactly, is the edge that agentic RAG has over regular RAG? @DattijoMuhammad sets out to find out through extensive testing — which produced counterintuitive results. https://t.co/DLAKDZmemR

[AI_Explorer @ai_explorer25]

Best 15 accounts to follow in AI:

@karpathy = LLMs king
@steipete = built openclaw
@gregisenberg = startup ideas king
@rileybrown= vibecode king
@jackfriks = solo apps king
@levelsio= startups king
@marclou = startups king
@EXM7777 = AI ops + systems king
@eptwts = AI money twitter king
@ai_explorer25= AI queen
@godofprompt =prompt king
@vasuman = AI agents king
@AmirMushich = AI ads king
@0xROAS = AI UGCs king
@egeberkina = AI images king

Follow them all and learn.

[Towards Data Science @TDataScience]

How well do AI image judges perform, and what are the areas where they still struggle?

@DattijoMuhammad explores a complex set of challenges in his new deep dive on vision AI models. https://t.co/ky6GCZ6bZ2

[Muhammad Abdullahi dattijo @DattijoMuhammad]

Caveats: small test, artificial differences, Gemini 3.1 Flash Lite instead of the main Gemini model (quota), and not comparable to VisionQ scores. Full article: https://t.co/ac1j3YCKzN

[Muhammad Abdullahi dattijo @DattijoMuhammad]

e. None of the three did. But they walked into a different trap. Thread:

[Muhammad Abdullahi dattijo @DattijoMuhammad]

GPT-4o fell for a different one. On the tilt question, where neither image was tilted more and the right answer was SAME, it picked an image 72.5% of the time, often sounding confident. Claude did that 17.5% of the time, Gemini 28.3%.

[Muhammad Abdullahi dattijo @DattijoMuhammad]

Retrieval completed 88.9% of tasks on mini and 85.6% on Haiku, and fetched every needed piece in 93% of runs. Caveat: 30 tasks, one made-up person, so read it as what to test, not a verdict. Full article: https://t.co/58EzBJcwen

[Muhammad Abdullahi dattijo @DattijoMuhammad]

I tested four ways to give an AI assistant a memory. The obvious one, pasting the whole history into every prompt, was the worst value: 20 to 31 times the cost of retrieval, and it did not win on accuracy either. Thread:

[Muhammad Abdullahi dattijo @DattijoMuhammad]

Full history also struggled with corrected facts: mini completed 46% of tasks on corrected facts vs 66% on unchanged ones. The running summary lost facts too: mini kept the newest value for 20 of 30, Haiku for 28.

[Muhammad Abdullahi dattijo @DattijoMuhammad]

Final setup: a cheaper model with the strong model as backup. 129 calls, $0.008, all 101 matches, zero false positives. Caveats: one run per setup, and I picked the escalation threshold after seeing results. Full write-up: https://t.co/t9y9RWoOCY

[Muhammad Abdullahi dattijo @DattijoMuhammad]

I made an apartment search agent 25x cheaper and it still found all 101 true matches with zero false positives. Cost for 500 listings x 5 requirements went from $0.20 to $0.008. Sending every pair to the strong model would have cost $1.88. Here is what actually moved the bill.

[Muhammad Abdullahi dattijo @DattijoMuhammad]

Plot twist: giving the strong model LESS text helped. On one hard listing, the full listing prompt got it right 0 of 10 times. Title and body only got it right 10 of 10.