Bryan Lim πΈπ¬ β
retweeted
Regarding the Wes Huff situation:
1. Christians, no matter their notoriety, should always be held accountable if they are truly found out to have sinned by plagiarizing. It's lying & theft.
2. Far too many Christians are using this situation to grow their own platform, smelling blood in the water. Some have a good heart, but I believe that to be far & few between.
3. It's profitable to wait to speak until more information is available, instead of defending or destroying right away.
4. Dan McClellan, many atheists, and enemies of Huff, who have had him in their crosshairs since his Billy Carson debate, are using this to destroy him because they serve Satan. It's not about proper Biblical accountability to them, though some of what they say may have merit.
5. Pray that the truth is revealed as Huff has seemingly been open to a proper & full investigation, while submitting to the elders of his church and leaders to the ministry he's involved in. Also, please be praying for his sick daughter.
God bless
Bryan Lim πΈπ¬ β
retweeted
My friend Robert Rouse (@bibleviz) spent years building open source data connecting the BIble's people, places, and things, and now he has compiled a list of great tools using that data. How cool it is when Christian share! viz.bible/theographic-s-impaβ¦
Bryan Lim πΈπ¬ β
retweeted
Our newest thing.
Introducing: anfra, an open-source engine for trusted analytics with AI agents.
Your coding agent builds the app in HTML, JS and any chart library. anfra connects it to governed metrics, with drill-downs, comparisons and lineage built in.
anfra.ai
Introducing: anfra, an open-source engine for trusted analytics with AI agents.
Your coding agent builds the app in HTML, JS and any chart library. anfra connects it to governed metrics, with drill-downs, comparisons and lineage built in.
anfra.ai
Saw on my timeline but can't find @levelsio post on ssh to his server.
I think that's the future. Where bots will ssh around an api/cli and make decision or purchases faster than human does.
11
Bryan Lim πΈπ¬ β
retweeted
She literally revealed the Abraham Hicks technique that stops overthinking fast βΌοΈβΌοΈ
it's true that we will look back at history and say
What?! we actually think that human can write better code last time? crazyyy
24
There are so many things codex and claude can do, but we are stuck with debating whether it is right or wrong to use it
oh well. It's just a tool
1
60
The cure for this world is not money or power. Or even right from wrong.
It's thankfulness, gratitude, appreciation, contentment, courage, faith, hope, LOVE
God is LOVE
16
youtube.com/watch?v=VstgBqdx⦠what an inspiration to be doing this to help the world! #james_tour $94 dollar to replace those 80K machines
17
I think chat.com or any LLM provider, should give us a link that open a new chat in temporary and sidebar closed. This will be useful for presentation and demo while keeping our privacy
28
Mmm!
ok it's official... AI-written unit tests and integration tests are empirically proven to be unhelpful. you should tell your agents to stop writing tests by themselves
on the deepswe eval set, banning sonnet 5.5 high from writing any tests actually resulted in slightly higher success rate (non stat-sig), with less time and token spent (stat-sig). this is pretty hard empirical evidence
across the tests that were written by the "tests allowed" baseline arm, 65% of them were unit tests, 35% were integration tests, neither bucket resulted in any improvement compared to not writing any tests at all
i also picked a random sample of 44 tasks subset where i completely disabled executing even existing tests - it also did not affect success rate at all
i spot checked many tests written in the baseline arm, and my intuition is that most tests are simply a repetition of the implementation
agent-written tests do not add any value because both the implementation and the tests were simply the agent's interpretation of our intent. the tests aren't any more accurate than the implementation itself
i suspect we can still extract some value from unit tests and integration tests if we describe them ourselves when we believe we can articulate our intent better through test cases than through requirements. but i have not proven this yet
also worth noting, during deepswe eval the agent wrote almost no e2e tests (only 17, compared to 3000+ unit/integration tests written). so this analysis does not prove nor disprove the value of e2e tests - i will do another eval specifically for that
33
As the calculus changes. Maybe we can see a fourth vertical on top of web, android, ios, that is made up of not html but native apps that are open.
1
29
Bryan Lim πΈπ¬ β
retweeted
Honestly... I was pretty against running multiple agents at once, but if you run `/grill-with-docs` from Matt Pocock and you get in full alignment with the agent, it'll spin up some really good work















