GPT 4.5 - Hit or Flop? Some Interesting Use Cases

Alexa gets smarter with Claude, GPT-4.5 becomes more human, and ElevenLabs Scribe sets a new standard for speech-to-text.

Author

Daily AI Skills
March 02, 2025

Welcome back to Daily AI Skills.

Here’s what we are covering today:
1. OpenAI GPT-4.5 : The Most Knowledgable Model
2. Amazon Alexa+ - Anthropic collaborates with Amazon
3. ElevenLabs Scribe V1 - The most capable speech to text model

GPT 4.5: The Most Knowledgeable and Emotionally Intelligent Model

OpenAI has unveiled GPT-4.5 (code-named Orion), its most advanced model yet, designed to enhance conversational depth and emotional intelligence. Here’s what you need to know:

  • More natural conversations – Uses unsupervised learning to improve world knowledge and intent recognition. It’s OpenAI’s last non-reasoning model.

  • Fewer hallucinations, better accuracy – Stronger across pro tasks, creative work, and general queries. It scores 62.5% on the SimpleQA Accuracy Test (o3-mini scored 15%) and scored 37.1 % on SimpleQA Hallucination Test (o3 mini scored 80.3%)

  • No major improvements in math or science – But surpasses o3-mini and o1 on SWE-Lancer, OpenAI’s freelance coding benchmark.

  • Limited access at launch – Available now for Pro users and paid developers, with Plus and Team users gaining access next week.

  • Sky-high API pricing – Costs $75 per million input tokens ($150 per million output), a 30x markup over GPT-4o.

The model doesn’t seem to be an upgrade to normal people, but if you’re a “high taste tester” - you should easily be able to detect the difference between GPT 4 and 4.5. Some interesting use cases have also arisen. The model is great at writing and generating SVGs and is amazing at medical image diagnosis.

Amazon Unveils Alexa+: A Smarter AI Assistant with Claude in the background

Amazon has introduced Alexa+, a next-generation AI assistant powered by generative AI, designed to be more conversational, intelligent, and action-oriented.

Key Features:

  • More Natural Conversations – Alexa+ understands casual speech, incomplete thoughts, and complex requests, making interactions feel more intuitive and human-like.

  • Takes Action Beyond Responses – Unlike traditional voice assistants, Alexa+ doesn’t just provide information—it can book reservations, order groceries, manage smart home devices, and even schedule repairs on its own.

  • Highly Personalized Experience – Alexa+ remembers user preferences, past purchases, and even details like dietary restrictions or important dates, allowing for customized recommendations and assistance.

  • Expanded Device Compatibility – Available on Echo devices, a new mobile app, and a browser-based platform at Alexa.com, with seamless conversation continuity across devices.

  • Agentic Capabilities – Alexa+ can navigate the web to complete complex tasks without supervision, such as finding and booking a service provider for home repairs.

It will cost $19.99 monthly but comes free with Amazon Prime membership, with early access rolling out in the U.S. next month. It is said to be using Anthropic’s Claude and Amazon’s Nova as its LLMs

Eleven Labs Scribe V1 - The best speech to text model

ElevenLabs has introduced Scribe, a cutting-edge speech-to-text model that claims to be the most accurate in the industry, surpassing competitors like Google’s Gemini 2.0 Flash and OpenAI’s Whisper v3 across multiple languages.

Key Features & Advancements

  • Unmatched Accuracy – Scribe boasts over 95% accuracy in more than 25 languages, including English, Spanish, and Italian, while supporting a total of 99 languages.

  • Enhanced Support for Low-Resource Languages – It significantly improves transcription for languages that have historically lacked reliable speech recognition, such as Serbian, Cantonese, and Malayalam.

  • Advanced Functionality – The model includes multi-speaker labeling, word-level timestamps, and even detects non-verbal audio markers like laughter and music.

  • Cost & Availability – Pricing starts at $0.40 per hour for pre-recorded audio transcription, with a low-latency version for real-time applications on the way.

As AI gets sharper, chattier, and more proactive, we're one step closer to a future where your virtual assistant knows what you need before you do—just hope it doesn’t start charging you rent.

📩 Forward it to people you know are keeping pace with the changing AI world and stay tuned for the next edition to stay ahead of the curve!