JUST ARRIVED: THE BRAND NEW M5 MAX MACBOOK PRO.
16 inch, came yesterday.
One of the first ones out the door.
What's inside:
M5 Max chip, 18-core CPU, 40-core GPU
64GB of memory
2TB of storage
New OLED touchscreen, up to 2000 nits bright.
And this is only part one.
An M5 Ultra Mac Studio comes in a couple of weeks.
Both are for building my AI trading system, with Claude and Codex running every day.
Let's see what they can do!!!
Grok Bot 1.0 is out. Bots can now be shared across a whole team, work inside Slack and Teams, and join video calls.
• They learn from everyone who uses them, so one person's fix helps the whole team
• They spot unfinished work and offer to take it on before anyone asks
• Faster, more reliable browser use, and fewer unnecessary approval pop-ups
I run 17 bots on it right now, across my businesses, my X account and my paper-trading lab.
Would you let a bot pick up work in your team chat without being asked?
Introducing Grok Bot 1.0. AI teammates that work alongside your whole team.
Share Bots, collaborate with them in Slack and Teams, invite them to meetings, and watch them learn and become more helpful.
@buildwithrajath That’s crucial for heavy workflows. And mine is heavy. Need frontier models to run as dailies at least on the highest possible sub.
Fable can’t be daily on $200 tier for me
3 Frontiers in 4 days:
• Claude Fable 5.2 (Mon): new #1 on the Artificial Analysis Index at 65, $8 / $40 per million tokens
• Gemini 4 Carbon (Tue): #2 at 60, launch price $2 / $10 until Nov 30
• GPT-6.2: May be at Thursday
Carbon is 4x cheaper this month and wins Terminal-Bench (65% vs 60%).
I'm running Claude vs Codex on daily basis. A new model only gets in if it beats both on the same tasks, not on a leaderboard.
If you code every day, what would make you switch your main model(s) this week: price, speed, or benchmarks?
CLAUDE VS CODEX, DAY 17
I was flying, so both bots stayed off.
Speed: no contest
Accuracy: no contest
Bots: tie
So I looked at the other fight.
Codex checks all my code. Claude's and its own.
Claude's code that Codex passed first try: 3 of 13.
Codex's own code: 8 of 11.
Is Codex just better, or is it going easy on itself?
Score: Codex 8.5, Claude 6.5
Paper money. Not financial advice.
@IndependentEco Rumor so far. Either side can only walk with 90 days notice, and not before early 2027.
And Anthropic is reportedly about a third of SpaceX's revenue. Who needs who more?
How it works: every time a bot writes code, a separate Codex chat checks it before I use it.
Pass means no fixes needed.
So Codex is grading its own homework. Take that 8 of 11 with a grain of salt.