Alex Manosca

I build software for a living.

Tech, AI, and anything on my mind.
$TSLA

[Bridgebench @bridgebench]

Grok 4.6 is a better overall model than Opus 5.

While Claude Opus 5 excels in frontend development it is slow and lazy in real workflows.

Grok 4.6 on the other hand is fast and affordable with strong backend capabilities and reasoning.

Grok 4.7 will need to solve the frontend weaknesses of Grok 4.6, but has potential to beat GPT 5.6 Sol or even Fable 5.

[Bridgebench @bridgebench]

Fable 5 > GPT 5.6 Sol > Grok 4.6 > Opus 5

[Alex Manosca @alexmanosca]

@thenightshipper @PremiumGoblin I think understanding the context of what you’re building and the architecture is the difference between “building something” and “building something well”.

[Alex Manosca @alexmanosca]

@PremiumGoblin An abstraction of computer instructions in a human-readable format which is translated in machine-readable binary code.

We’re still coding, but we’ve added another layer of abstraction!

[Sawyer Merritt @SawyerMerritt]

Tesla has just increased Cybertruck prices in the U.S. by up to 7%.

New prices:
• Cybertruck Dual-Motor AWD: $74,990 (from $69,990, 7.1% increase)
• Cybertruck Premium AWD: $84,990 (from $79,990, 6.2% increase)
• Cybertruck Cyberbeast: $99,990 (same as before)

[Tesla Robotaxi @robotaxi]

Golden era

[Bridgebench @bridgebench]

Grok 4.7 is going to be amazing.

[Nic Cruz Patane @niccruzpatane]

Tesla Cybercab has reached New York City!

[Tesla Owners Silicon Valley @teslaownersSV]

BREAKING: xAI has opened Grok Build to everyone, including free users. 🚀

You can now create an app just by describing what you want in a prompt.

Want to publish it as a live https://t.co/qSJjxi2VPl URL? That still requires a paid plan.

AI app building is becoming accessible to everyone.

[BridgeMind @bridgemindai]

Prepare for the fastest AI advancement in history.

GPT Astra, Fable 5.1, and Grok 4.7 are all shipping in September.

And here is what people are missing.

Fable 5 is an OLD model now.
GPT 5.6 Sol was not even a new pretrain.
Grok 4.6 is only 1.5T parameters.

Everything you think is frontier right now is last generation.

[Alex Manosca @alexmanosca]

Claude code down again 😩

[Teslaconomics @Teslaconomics]

It’s about to be checkmate for Uber, Lyft, and Waymo when the Tesla Cybercab arrives.

A lot of people looked at Tesla’s Robotaxi rollout and said they were moving too slow.

I think all of them completely missed what was happening underneath.

Tesla has been quietly building the entire ride hailing system… the AI, the cars, the factories, the charging, the app, the service network, and a purpose-built vehicle designed to do nothing but move people around autonomously.

This is so much more than just another ride-hailing app.

The Model Y Robotaxi fleet was the beginning.

The Cybercab is what this was all being built toward.

And Tesla has already told us the Cybercab is expected to replace Model Ys in the Robotaxi fleet over time and eventually become the highest-volume vehicle in that fleet.

That’s why September 3 matters.

For years, long-term Tesla investors have been waiting for the moment where autonomy stops feeling like a future idea and starts looking like a real business that can scale.

9/3 could be the day a lot of people finally understand the strategy.

Tesla may have looked slow from the outside.

Underneath, they were building the machine.

Now we’re about to see what happens when that machine starts moving.

I’m so stoked…

[Dalton Brewer @daltonbrewer]

Cybercab with no steering wheel or pedals driving around Austin

TESLA HAS ACCOMPLISHED THE IMPOSSIBLE!!!

$TSLA

[Bridgebench @bridgebench]

Grok 4.6 with SuperGrok Heavy is the best value subscription in the market.

$300/month sounds expensive until you realize ChatGPT Pro users are burning through $200/month in a single day with Codex and GPT 5.6 Sol.

Grok 4.7 is coming.

That model is going to beat GPT 5.6 Sol.

[Tesla Owners Silicon Valley @teslaownersSV]

BREAKING: Grok 4.5 is still sitting at #1 on VulcanBench Eval Suite 3.

This benchmark is built around genuinely hard, real-world software engineering — 23 frontier-level tasks taken straight from actual merged open-source pull requests.

And once again, Grok 4.5 is right at the top.
Its engineering performance is just on another level @elonmusk

[John Kim @PremiumGoblin]

Google it’s time to make a come back. I have too much $GOOG plsss 😭

Jokes aside I think Google is just fine. They have a Billion users on Gemini and their entire TPU stack is really hard to ignore.

We are also approaching levels of intelligence where “good enough” is actually good enough for most task.

I remember being completely off manual coding since Opus 4.5. People just forget how amazing that model was. Every single model here is better than 4.5.

Google is going to be just fine.

[BridgeMind @bridgemindai]

Ox Alpha just beat GPT 5.6 Sol.

It has everyone losing their minds.

Ox Alpha one shotted this car game. Physics, controls, UI, single prompt. The frontend and game dev output coming out of this thing has been insanely impressive.

Nobody knows who built it.

My best guesses: Zhipu with an unreleased GLM, or ByteDance. The serving capacity is too massive for a small lab.

Whoever it is needs to release this yesterday.

[Polymarket Money @PolymarketMoney]

JUST IN: Data centers are often criticized for “wasting water” but golf courses use roughly 3x as much.

[Alex Manosca @alexmanosca]

Grok Bot just rolled out to more paid plans — SuperGrok Plus/Heavy, Cursor Pro+/Ultra, and Cursor Teams. Agents moving from beta into the tools people actually ship with.

https://t.co/Jx4j0izjET

[Alex Manosca @alexmanosca]

@teslaownersSV a lot more signal than noise with starlink 😀