Beutifly written, pragmatic approach what defense in depth means in ai era, still relaying on decades old approaches with taking advantage of modern tech substack.com/home/post/p-223…
59
Quite interesting Google joining typesafe.ai JAV game
Today we’re launching the Gemini Decisions API, powered by our new Gemini Decision 1 model in @GoogleAIStudio.
Instead of generating text token by token, it’s built for closed-set choices scoring up to 1,000 questions across text, images, audio, video + PDFs in a single pass. A few details:
- 3 decision types: yes/no, multiple choice (up to 255 options) + rubric scoring
- $0.10 / 1M input tokens with $0 billed for output tokens (~5x cheaper than Flash-Lite)
- leads on several benchmark evals with lowest calibration error
Google teams are already building with it, below is a look at Stitch Rapid designing a full product flow in seconds (1x speed)
58
Turn ON a Claude and turn OFF brain is unfortunately becoming a de-facto standard. No, it doesn't have anything with productivity. It does with critical thinking, added value. Burning tokens doesn't remove a need for engineering skills ( no skilld.md reffered)
47
True cost of bad hire: ruined team morale as the worst price payed. Problems with ownership, draining team energy constantly, ...
2
122
Toto pobavilo a děje se to všude :-D Koncept "dočasně trvalý" #swengineering #programming centrum.cz/clanek/programato…
45
3 weeks of Digital detox clears mind, returns ability to see things clearly, name It correctly ... Now trying to preserve it as long as possible
2
136
As #tech lover I keep an eye on #ai and #cybersecurity. Was building a #RAG based system before it was fancy to name it lake that.Have opinion, went through at least 2 hype cycles. @kelsayhightower coined it as #zte zero token architecture youtu.be/A7WFt2JQ5sg?is=usqQ…
well pronounced
137
That was quick! 1 day after #anthropic mythos And fable release 120k system prompt leaked! They Are the same in Essence. #security doesnt mean you ask model kindly not to do sth. #ai #llm cybersecuritynews.com/anthro…
69
Interakce s @vodafone eshopem je opravdu za trest. Support trvá 2 pracovní dny a 30s na to se dostanete opět do stejného nepoužitelného stavu.
1
52
Nice interactive #ai #llm Pareto frontier based on llm arena frontier model comparison michaelshi.me/pareto/
48
When I was working mainly with #gcp cloud I couldn't have understood how @datadoghq, @splunk, and other services could make their money.Untill I went back to #aws cloud at least partially.Than I understood. Small diff in product management approach, feature driven Vs value driven
57
Just if you missex that Peter Norvig experiment on 2025 Advent of Code #AoC #LLMs for Vibe coding github.com/norvig/pytudes/bl…
Antivirus becoming useless. Now using AI LLMs generating code on the fly to bypass traditional detections systems #CyberSecurity
blog.google/technology/safet…
56
Completed Go To Market strategist by @majavoje. It Is so packed! Ton of resources and references. Great insights from sharing real insights.
- Beach head strategy vs Spray And Prey
- how to build growth loops
... And more. Thanks for broadening horizons!
39



