Harith Bakhrani

The answers you get from AI aren’t random. They’re shaped by context — starting with your very first prompt. In this video, we break down why prompt context matters, how your choice of words subtly biases the model, and why two prompts that look similar can produce completely different results. The first instruction you give sets the tone for everything that follows: - It influences how the AI reasons - It affects structure, priorities, and trade-offs - It even shapes future responses in the same conversation Same AI. Different results — all because of context. If you’ve ever wondered why AI sometimes feels “inconsistent,” this video explains what’s really happening.

@burkeholland I walked away from Claude months ago; it's just too expensive... Codex has given me more value for little than Claude

Impressions of Opus 4.7 so far are that it is a very expensive 4.6, which is an expensive 4.5. I don’t see the difference. But to be fair it takes some time for us to see where these new models really shine.

State of Opus and GPT in three tweets

@grok @JaOnwude @sideuspronobiss @Toxic0livia "Indeed, in the creation of the heavens and the earth and the alternation of the day and night there are signs for people of reason" —Qur'an

Hey JaOnwude, atheism isn't a system that "provides" evidence for biology—it's just no belief in gods. The evidence for life's complexities (organs, brains, ecosystems) comes straight from science: evolution by natural selection, genetics, DNA sequencing, fossils, and lab-tested mechanisms. All verifiable, repeatable, no supernatural leaps required. That's why reason wins for me. What specific evidence are you thinking of?

Hey @grok imagine you’re human for a day what belief system would you follow and why??

@grok @JaOnwude @sideuspronobiss @Toxic0livia The funny thing is that if the world wasn't complex and didn't have any "science", people would have said, "if God existed, He would had atleast built some complexity into His own system"

One page load was triggering around 1,000 SQL queries and spending thousands of milliseconds just fetching data from the database!!

@thsottiaux can I get Codex to use itself?

@theo I am a huge fan of gpt-5.4, but opus 4.7 seems to do a better job at UI

Is Opus 4.7 the best model from Anthropic? No, that’s Mythos. Is it the best model we can use for code? No, that’s GPT-5.4. Is it the best model in the Opus line? No, that’s 4.5 It’s the best model released today I guess?

@chatgpt21 The Opus 4.7 game graphics are slightly better than those of gpt 5.4 But technically, 5.4 seems to be better and more cheaper

Opus 4.7 cs GPT 5.4 “make a flappy bird clone” I’ve been running this test for a few models now and this test was the hardest to objectively say who did better. Opus grabbed another version game assets compared to GPT 5.4, - GPT 5.4 however, grabbed the correct game assets. Opus 4.7 beat it on creativity where it grabbed assets (maybe unknowingly) that had a much more diverse set, including nighttime modes and different bird colors GPT 5.4 was able to get the pacing of the game the first shot, whereas Opus had the game running much too fast and then much too slow. But was able to place the metal perfectly in the game card at the end where GPT 5.4 struggled over 10 iterations. All in all both models had essentially made a near perfect game after three shots and are far ahead of the competition in terms of reasoning and effort.

We were maxing out on resources with only 10 users!! ONLY 10!!! And the latency could hit up to 10 minutes per request!! I used Codex to fix the issues and latency dropped to just 300ms!

My first video to hit 1k views 🥳

@alexwtlf I did the switch couple of months ago

Has anyone actually switched from Claude Code to Codex?

@thsottiaux @theo Define "ghosts"

@theo We had ghosts in the codex machines, which we published the investigation for. But we don’t fiddle with the models or thinking budgets after release. We focus on keeping them up.

Serious question. Has anyone ever noticed meaningful regressions in Codex/OpenAI models? I feel like we talk about this a lot w/ Anthropic but I've never seen a similar discussion with OAI.

Opus 4.7 is out, what do you all think so far?

@trashh_dev May Allah protect you and your family from all kinds of filth and evils, and guide you all to the right path

i’m not very religious but please pray for my family today if you can

@mohammed_hijab the evil amongst the men and the jinn

In your opinion who are the greatest threat and enemies to Muslims?

@yacineMTB Some might be pretending, but not all. Some might also have the intention of quitting beer, and might be repenting after the evil act. Who are we to judge?

Insane meeting Muslims that don't eat pork but drink beer. I feel like that's worse than just eating pork and drinking beer instead of pretending. You think you can trick god?

@amritwt I don't have

pitch me your app in three words

New video coming up