Chris Kukelhan

Owner - Highgarden Real Estate. Love building a brokerage that agents can be proud of. Local | Culture | Support | Marketing | Tech | Growth | Independent

[Chris Kukelhan @ChrisKukelhan]

@DerekFeehrer windows please :)

[Logan Kilpatrick @OfficialLoganK]

@llm_wizard putting the means of software creations into the hands of everyone is going to be one if the most economically transformative things in history

[Sundar Pichai @sundarpichai]

1/ Just delivered Q3 earnings remarks. A few additional highlights from the call:

Our AI Models, Gemini 2.5 Pro, Veo, Genie 3 + Nano 🍌 are leading the way. 13M+ developers have built with our generative models. Looking forward to the Gemini 3 release later this year!

[Google AI Studio @GoogleAIStudio]

we heard you, saved system instructions are here!

now you can create, save, and reuse them across chats

[Chris Kukelhan @ChrisKukelhan]

@GoogleAIStudio Nice. This is a huge UX upgrade. A small thing with a big impact. Thanks for not taking your eye off the user.

[Chris Kukelhan @ChrisKukelhan]

@OfficialLoganK @GoogleAIStudio Thank you. A little thing, that will save a ton of time. :)

[Chris Kukelhan @ChrisKukelhan]

Why is there no option to connect to Supabase any longer? I just saved and then forked a project in Stackblitz, loaded in https://t.co/bp9LwTCTML and there is now no preview displayed and no way to link my project back to the Supabase database that I have used on this project for the last 6 months. Please help. Thank you. @boltdotnew @boltdotnewhelp @supabase

[Chris Kukelhan @ChrisKukelhan]

@levelsio @OpenAI I wired it up in an quick app and using Sora-2 the quality seemed not great but there wasn’t a watermark. You using Sora-2-pro? Hi res?

[Chris Kukelhan @ChrisKukelhan]

@levelsio @Lost_Signal There’s an opportunity here. Not easy, but doable.

[Chris Kukelhan @ChrisKukelhan]

@sebastianvolkis I’ve had good luck with Gemini when I turn the temp way down. Like .2

[Chris Kukelhan @ChrisKukelhan]

@chamath

[Chris Kukelhan @ChrisKukelhan]

@novapilgrim

[Chris Kukelhan @ChrisKukelhan]

Looks like one person got the message. Lol. #returntowork

[Mustafa Suleyman @mustafasuleyman]

After Ethan's post, I went on a deep dive into this study! I could go on and on about the results but if I had to boil it down to my biggest takeaways...🧵

[SpaceX @SpaceX]

Mechazilla has caught the Super Heavy booster!

[Chris Kukelhan @ChrisKukelhan]

Works perfectly.

[Chris Kukelhan @ChrisKukelhan]

@Jason

[CNN @CNN]

President Biden decried the apparent assassination attempt on Donald Trump, saying the Secret Service is working to see if "further adjustments" are needed to ensure his safety https://t.co/ZJho2qbOp0

[Chris Kukelhan @ChrisKukelhan]

Found this.
Controlling costs
To manage costs with the o1 series models, you can limit the total number of tokens the model generates (including both reasoning and completion tokens) by using the max_completion_tokens parameter.

In previous models, the max_tokens parameter controlled both the number of tokens generated and the number of tokens visible to the user, which were always equal. However, with the o1 series, the total tokens generated can exceed the number of visible tokens due to the internal reasoning tokens.

Because some applications might rely on max_tokens matching the number of tokens received from the API, the o1 series introduces max_completion_tokens to explicitly control the total number of tokens generated by the model, including both reasoning and visible completion tokens. This explicit opt-in ensures no existing applications break when using the new models. The max_tokens parameter continues to function as before for all previous models.

https://t.co/yj5FGGBwg2?

[Chris Kukelhan @ChrisKukelhan]

@OfficialLoganK I am seeing solid improvements in image analysis. Gemini 1.5 Flash Experimental 0827