Jacob Shinas

I'm not looking for people to tell me it's great. I'm looking for honest feedback about what's confusing, broken, unnecessary, or missing. Join TestFlight here: testflight.apple.com/join/fVAJbbbv Please note this is iOS only for now.

@forgebitz Time to use Cursor

grok is stealing your code gpt is deleting your code claude is taking your money

@DinkinFlicka400 @techdevnotes Sounds about right. On the coding front I agree it’s not on the frontier but the cost to performance makes it too good to pass up for me. Hopefully Grok 5 bridges the gaps.

For me coding still sucks and image creation(for example people are heavily praising it's ability for media creation) compared to Google or OpenAi or ByteDance etc... It is really bad The only part I like Grok and I really think it is above competition is "Live searching" where you ask it question and it will live search latest stuff and give you correct results

We are still waiting for Grok Desktop

@DinkinFlicka400 @techdevnotes Hasn’t been my experience using it but I guess it could depend on the use case. I’m curious as to why you think it’s bad though.

@JacobShintas @techdevnotes No offense but 4.5 is still shit, just bunch of internal benchmarks and Elon PR team on twitter hyping it

@DinkinFlicka400 @techdevnotes If you’re still waiting after 4.5 you’ll be waiting forever

@techdevnotes I am still waiting on Grok model that is not complete shit

@GoogleAIStudio Gemini 3.5 Pro

What are you vibe coding this weekend?

@mark_k @AnthropicAI Shocking

Fable News: @AnthropicAI is extending Claude Fable 5 access across all paid plans through July 19. Claude Code’s weekly rate limits will also remain 50% higher until then.

@Erdayastronaut @OpenSauceLive @RocketRanchTX In the future people will start predicting the launches based off your traveling.

Hoping they nail all the objectives!!! Per Dodd’s law I’m flying to @OpenSauceLive on Thursday 🫠 so we’ll see what our streaming situation will be 😪 I just want to see it in person at @RocketRanchTX 😭

@wholemars Kinda funny how he doesn’t have a snarky comment on this post

March 11th, news of the deaths comes out March 18, Zack Nelson (JerryRigEverything) takes money from Ford to push this unsafe technology on his viewers and starts attacking Tesla Self-Driving that is good enough to save lives. He is willing to kill his viewers for money.

She can drive, she just gets tired. Not having to drive herself makes her less tired. What kills people is your sponsored videos promoting inferior technologies like Ford Blue Cruise. Remember how you took money from them to help cover up two deaths? @ZacksJerryRig?

@SurfingPigment @LexnLin Bro no

@LexnLin 3.1 pro is still better than 4.8… I don’t understand why people don’t know how to use gemini

If DeepMind doesn't release Gemini 3.5 Pro this month, well then we will get Opus 5 sooner or later and in 1.5 months probably GPT 6 How tf do you want them to compete at all? I just hope Gemini 3.5 Pro will be Opus 4.8 level.

@pawzzard @ns123abc He’s Megatron IRL frfr

@ns123abc bro has been going to jail for like 3 years now altman got more lives than a cat at this point

openai is going to disappear and scam defaultman is going to jail 😭 it’s over

@kvickart Cursor Pro+

grok needs a middle tier

@JonDevX @shubgaur It’s MOE which means only a portion of the model’s size is used for each task.

@shubgaur Does it mean the model is small? 😐 I'm very very excited about this release, 6$ out is overwhelmingly lower than what I would expect.

Leaving this here for the valuemaxxers... Cursor Grok 4.5: $2 in / $6 out GPT 5.6: $5 in / $30 out Opus 4.8: $5 in / $25 out Have fun

@theo Please do

Grok 4.5 is good enough that I think I have to do a video on it...

@ArtificialAnlys That's an insane jump

SpaceXAI just released Grok 4.5, and it ranks #4 on GDPval-AA v2 with an Elo of 1543 - behind only the latest Claude releases from Anthropic on real-world agentic knowledge work tasks Grok 4.5 achieved this score at a cost of $0.49 per GDPval task to sit clearly on the Pareto frontier for performance versus cost. This cost is lower than GLM-5.2 and Kimi K2.6, and nearly 90% cheaper than the models ahead of it on our leaderboard. We’re finalizing the remaining Artificial Analysis Intelligence Index evaluations and will share final results soon. Thanks to @SpaceXAI and @elonmusk for their collaboration testing this model ahead of release, and congratulations on the launch!

@cursor_ai LETS GO

Grok 4.5 and Composer are two different model weight classes. Composer 2.5 will remain offered, and we will release new models of this size going forward.

We've partnered with SpaceXAI to train Grok 4.5. It’s our most powerful model yet and the first we've built for more than software engineering.

Grok 4.5 pricing is disruptive if it performs.

@Coolshit4you @ravikiran_dev7 Just too much content and too many creators already on YouTube.

@ravikiran_dev7 Why buy something that you could incorporate a better version of into something you already own? XTube could live right here!

Unpopular opinion: Elon musk should buy YouTube.

@ns123abc I really hope this model is good. Cannot wait to use it.

BRO LITERALLY PREDICTED THIS

@theskory AYEEEE

Today!

@elonmusk @SpaceXAI Well I know what I’ll be doing for the next 12 hours *refreshes Grok

Based on strong positive feedback from customers in our beta test program, @SpaceXAI will make Grok 4.5 available to the public tomorrow. It is an Opus-class model, but faster, more token-efficient and lower cost.