John Logan

@MSFT365Status Had an issue today where all of our rooms showed up in outlook as unavailable although you could still book timeslots. Looks like recent updates have highlighted mismatches in the WorkingHoursTimeZone between the outlook client and the room mailbox.

@GeminiApp I am going to put it out there. I think this release is flawed. When asking it to build software it adds too many features that you didn't ask for. It's like an assistant that's trying too hard to impress and not doing what it was told.

@GeminiApp Just used Gemini to build an app that guesses the regular expression required to extract fields from pasted text into a CSV file. No more having to work it out manually! When the usual Tools for PDF's tabular data extraction don't cut it, this is a game changer.

Poor nova has hurt her foot @Samshipley14

When I was a young boy I told my grandfather that one day you would be able to talk to a computer just like another person. He said that computers would never be as intelligent as humans. I said it didn't matter, they only had to reach the level that he wouldn't be able to know.

@📖 Bess the Naughty Puppy g.co/gemini/share/0…

@jacobhess @FeatureCrewPod In my mind the problem was always the way it was asked. Its trivial for the AI to come up with a method of solving a maze the issue was the way in which the question was presented. It was a test of visual processing rather than tracing a maze.

@Techomagician @FeatureCrewPod That’s what I was testing! If the lines were too thin, it seemed to get confused!

🚀 Excited to share our latest episode where Chris and Jakob dive into the full ChatGPT o3 model! 🌟 Join us as we explore coding, business reasoning, and agentic intelligence tests. Check it out! 👉 youtube.com/watch?v=D9QcBU… #AI #ChatGPT #TechTalk youtube.com/watch?v=D9QcBU…

@jacobhess @FeatureCrewPod then it allows the user to select a start point with the mouse and an end point. it then uses path finding algorithm to find the route from the start point to the end point without touching any of the walls. it display the route in a different colour

@jacobhess @FeatureCrewPod create a program in html and javascript in a single html file that will allow you to upload an image and then convert that image into a 1 bit file.

@jacobhess @FeatureCrewPod finally solved it, the solution was quite easy in the end. I had access to gemini diffusion which also meant the solution only took 2.7 seconds. I will put the prompt in my next reply

@leonardoai when will you fix flux kontext? wasted credits, don't publish models that don't work

@jacobhess @FeatureCrewPod Not sure I quite get the test?, I can solve using AI mazes of any size with a prompt. is it to solve it just from the image?

First, I love the pod. You are top on my goto list for each model release. Second, there was a comment on YouTube suggesting to only use the image, not the ascii, so, I gave it a go. It was an agonizing 9 minutes and 53 seconds, but it figured out the 20x20. (ChatGPT doesn’t share images, so, i attached them to this post!) chatgpt.com/share/68006b23…

@FeatureCrewPod I like your AI maze solve, I can do it using AI but I wanted to check that its not about "solving it" but "solving it with a simple prompt like - solve this maze"

@leonardoai I have long respected LeonardoAI since the beginning of AI art generation but after losing hundred of credits to failed generations in Flux Kontext (which clearly is bugged) I have cancelled my subscription.

@bindureddy I don't really trust any benchmarks on AI. I just get x,y and z to code a,b and c and compare the process and the result. Grok 3 has performed the best for me so far.

Grok-3-Thinking Scores Way Below o3-mini-high For Coding on LiveBench AI Grok-3 is a good model, and OpenAI bashers love Grok-3 thinking for obvious reasons. 😉 Objectively, however, it scores WAY BELOW o3-mini-high for coding, and it takes forever to answer the most basic coding questions. o3-mini-high - 82.74 grok-3 thinking - 67.38 On the plus side, it's a bit above Sonnet, though Sonnet is like 10-100x FASTER. TLDR: Grok-3 is good for great banter on X but isn't a great model for developers.

My first track on @SunoMusic V4, awesome AI music generator suno.com/song/8dae0401-…

I think I might be in love with Suno v4 suno.com/song/8dae0401-…

@Gizmodo Ha ha ha ha ha ha ha ha ha ha

Amazon's Best Tax and Security Duo, TurboTax Deluxe and McAfee Bundle Is 63% Off #deal gizmodo.com/amazons-best-t…