![Flux[dev]: Sam Altman in a grocery store holding a box of fresh strawberries. Next to him is a futuristic robot, sleek smooth humanoid design. Smooth, glossy black faceplate with no visible facial features, high-tech, minimalist appearance. The robot's body is matte black or dark gray, with articulated joints and mechanical parts that resemble those of a human, including fingers. In the background, "OpenAI" is written on a sign.](https://ethanbholland.com/wp-content/uploads/2024/09/openai-2.png)
Flux[dev]: Sam Altman in a grocery store holding a box of fresh strawberries. Next to him is a futuristic robot, sleek smooth humanoid design. Smooth, glossy black faceplate with no visible facial features, high-tech, minimalist appearance. The robot’s body is matte black or dark gray, with articulated joints and mechanical parts that resemble those of a human, including fingers. In the background, “OpenAI” is written on a sign.
Major shifts at OpenAI spark skepticism about impending AGI timelines | Ars Technica
OpenAI co-founder John Schulman says he will join rival Anthropic
“New: 3 OpenAI leaders are leaving. President Greg Brockman going on extended leave, John Shulman to rival Anthropic, and product leader Peter Deng is also out.
OpenAI co-founder Schulman leaves for Anthropic, Brockman takes extended leave | TechCrunch
“Guaranteed valid tool calls from OpenAI Tool calling is incredibly useful for building LLM-steered applications. One of the key difficulties with tool calling is getting models to reliably generate arguments that respect the tool schema. Earlier today OpenAI announced that you
“Ok, so this safety section explains some of the reason why GPT-4 voice mode was likely delayed & some novel risks of voice. Without guardrails, GPT-4 could imitate people (including the user), it could identify speakers & it would answer differently depending on how you spoke.
OpenAI Co-Founder John Schulman Departs for AI Rival Anthropic – Bloomberg
“We’re starting to roll out advanced Voice Mode to a small group of ChatGPT Plus users. Advanced Voice Mode offers more natural, real-time conversations, allows you to interrupt anytime, and senses and responds to your emotions.
“Use the best models for the use case Code – Sonnet 3.5 vastly superior to anything else out there Search / Information – GPT-4o Does a good job of summarizing information. Doc analysis and understanding – GPT-4 or GPT-4o Reasoning combined with good RAG capabilities” / X
“LiveBench Benchmarks are for the new GPT-4o-2024-08-06 Claude-3.5 still at no-1
“OpenAI GPT-4o mini is 15c per million input tokens, so this new Gemini Flash 1.5 pricing is half of that, at 7.5c/million I wonder if we’ll see a price drop for Claude 3 Haiku soon, that’s currently the most expensive of the main fast-cheap models at 25c per million for input” / X
“Long awaited, but finally – Structured Outputs. gpt-4o-2024-08-06 achieves 100% reliability in our evals, perfectly matching the output schemas. 🔥 Super useful for downstream task, when the model is not calling a tool, but rather, responding to the user in a structured way. A
“Structured output is all you need! Structured outputs can help to improve the performance and reliability of your LLM application. OpenAI just announced that they are now supporting structured outputs in the API. You can use it with function calling and defining your own
Introducing Structured Outputs in the API | OpenAI
OpenAI: Introducing Structured Outputs in the API
“On one hand, OpenAI just made it so that its models can be forced to output in JSON, which makes it much easier for developers to use LLMs in their programs. On the other hand, forcing LLMs to answer in JSON lowers their reasoning ability (but helps classification)” / X
“It went under the radar but OpenAI released their most performant GPT-4o assistant model yesterday. – Structured outputs with 100% reliability – 4x max token (16,384) output limit – 50% cheaper on inputs – 33% cheaper on outputs – #1 on the @allen_ai ‘s ZeroEval leaderboard
“OpenAI-backed Figure just revealed the world’s most advanced humanoid robot: Figure 02. This wild demo shows off the humanoid working in a BMW factory, completely autonomous with self-correcting behavior. The new technical features that stand out:
1. Speech-to-speech reasoning
2. Vision Language Model (VLM):
3. 20-hour workday
4. AI vision
5. Hands:
6. 3x computational power:
GPT-5 with the body is going to be quite the spectacle!”
OpenAI worries people may become emotionally reliant on its new ChatGPT voice mode | CNN Business
OpenAI Makes a $60 Million Hardware Startup Bet — The Information
OpenAI has built a text watermarking method to detect ChatGPT-written content — company has mulled its release over the past year | Tom’s Hardware
“Wait did OpenAI just drop the price of GPT-4o (the real one, not mini) by 50% without even announcing it? Like there is nothing on their blog/@OpenAIDevs , just a new pricing page with $2.50/1M tokens listed. Bullish for OpenAI. 🤯
ChatGPT’s mobile app just had its biggest month yet | TechCrunch
“Kind of a big deal: GPT-4 simulates people well enough to replicate social science experiments with high accuracy Note this is done by having the AI prompted to respond to survey questions as a person given random demographic characteristics & surveying thousands of “AI people”
How This Brain Implant Is Using ChatGPT – CNET
OpenAI Co-Founders Schulman and Brockman Step Back
OpenAI co-founder quits (again)
“I shared the following note with my OpenAI colleagues today: I’ve made the difficult decision to leave OpenAI. This choice stems from my desire to deepen my focus on AI alignment, and to start a new chapter of my career where I can return to hands-on technical work. I’ve decided” / X
(1) The Winds of AI Winter (Q2 Four Wars Recap) + ChatGPT Voice Mode Preview
“It is hard to overemphasize how fast AI adoption is happening. This is data on ChatGPT adoption in Denmark by job from January, 2024. 15% of journalists and marketers in had Plus subscriptions! We should be paying more attention to gaps in who uses AI.
How Does OpenAI Survive?
“My long-form review of ChatGPT’s Advanced Voice mode:
“accidentally found a really interesting ChatGPT prompt In a new chat ask “break this down into json format for me. be detailed and thorough” It’ll breakdown all the information it knows about you from what you’ve shared with it in previous interactions You can respond and
“@sama i kind of want openai to ship actual strawberries just as a bit” / X
How do people use ChatGPT? We analyzed real AI chatbot conversations – The Washington Post
GPT-4o System Card | OpenAI
OpenAI tempers expectations with less bombastic, GPT-5-less DevDay this fall | TechCrunch
SearchGPT is a prototype of new AI search features | OpenAI
GPT-5: Everything You Need to Know – by Alberto Romero
OpenAI invests in a webcam company turned AI startup. – The Verge
“OpenAI is leading a $60 million investment in Opal, which sells $300 professional webcams and plans to develop other types of devices powered by OpenAI’s AI models. w/ @steph_palazzolo
OpenAI says it’s taking a ‘deliberate approach’ to releasing tools that can detect writing from ChatGPT | TechCrunch
Figure’s new humanoid robot leverages OpenAI for natural speech conversations | TechCrunch
Leave a Reply