OpenAI Invests in Finance Agents

Good morning. It’s Friday, August fifteenth.

On this present day in tech history: In 2000MIT’s Cynthia Breazeal introduced Kismet, an expressive humanoid robot able to recognizing and mimicking emotional cues, an early, influential example of social robotics and affective AI. It fused real-time gaze tracking, prosody-aware speech parsing, and an ethology-inspired behavior engine to supply surprisingly lifelike mood shifts.

Q&A
OpenAI Invests in Finance Agents
Google Releases mini-model
Claude’s learning mode
5 Latest AI Tools
Latest AI Research Papers

You read. We listen. Tell us what you think that by replying to this email.

AI Breakfast Q&A

Every edition, we’ll answer real questions from AI Breakfast readers—practical, weird, and all the pieces in between. Got something you’ve been wondering about AI? In search of a particular tool? Hit “reply” to this email and ask! We may feature it in the following edition!

Pritty A: I head the chance management function and need to construct workflows like compiling KRIs from different departments with none IT background. What’s the simplest tool? To do that without code, start with Airtable or Notion. Each allow you to collect KRIs via easy forms, store them in a central table, and create dashboards. Pair them with Zapier or Make to routinely pull in data from emails, spreadsheets, or forms so your reports construct themselves. When you prefer a risk-specific platform, LogicGate Risk Cloud and Resolver are excellent (though pricier) options.

Ernie H: What’s a “token?” What does it consult with on the planet of AI? A token is only a small piece of text the model reads or writes. It may be an entire word, a part of a word, and even punctuation. Models process text as a series of those tokens, and limits like “200k tokens” Like in Claude Sonnet 4 mean the full variety of pieces of text it may possibly handle directly.

A.S. Ganesh: After I come across a job posting in LinkedIn/ company’s profession page, I would like AI to customize my resume as per the job description and apply on my behalf. Are you able to explain how I can achieve this? You may do that by combining just a few automation tools. Use Zapier or Make to capture the job description from LinkedIn or an organization’s careers page. Then feed that text, along along with your current resume, into whatever AI model you want (ChatGPT, Grok, etc.) to rewrite your resume so it highlights your most relevant skills, uses keywords from the job description, and matches the role’s tone. Once customized, automation services corresponding to LoopCV, LazyApply, or Simplify.jobs can upload your resume and auto-fill applications. Also browser-based tools like UiPath or Robocorp can handle LinkedIn submissions specifically. Good luck!

Reply to this email along with your query and it could get featured!

Today’s trending AI news stories

OpenAI juggles cost-cuts, recent monetization plays, and $14M bet on finance agents

Openai’s Startup Fund just injected $14M into Endex, a Thiel Fellow–founded startup integrating an AI agent directly into Excel. Powered by OpenAI’s reasoning stack, Endex can grind through multi-hour financial workflows: constructing money flow models, parsing SEC filings, converting PDFs to structured tables, all GDPR- and SOC 2–compliant. It’s aimed squarely at finance teams that need speed without losing audit trails, potentially stepping on Microsoft’s turf.

It’s official. We’ve raised $14m led by @Openai Startup Fund to bring AI to Excel.

Endex is the primary AI agent to live inside Excel.

For the past 12 months, we have been working with financial firms. Today we’re releasing it to the world.

Our capability is restricted; comment below for

– Tarun Amasa (@tarunamasa)
4:28 PM • Aug 6, 2025

Meanwhile, The Register calls GPT-5 “a value cutting exercise” saying the model is tuned for efficiency over brute force. As a substitute of 1 giant model, it routes easy queries to a “tiny” LLM and complicated ones to a heavyweight core. MXFP4 quantization slashes compute/memory by as much as 75% and reasoning is auto-throttled to chop token spend. Gains are modest other than an 80% drop in hallucinations, though GPT-5 showcased efficiency in style, completing Pokémon Red in only 6,470 steps versus o3’s 18,184.

OpenAI stays cautious on monetization strategy. Nonetheless, ChatGPT head Nick Turley says promoting is “not ruled out” but would require careful integration, while CEO Sam Altman frames it as a “last resort.” OpenAI can be piloting “Commerce in ChatGPT,” enabling purchases directly via chatbot recommendations without allowing affiliate revenue to skew results. Read more.

Google releases pint-size open-weight AI model, adds personalization and memory to Gemini

Google released Gemma 3 270M, an ultra-compact open model with just 270 million parameters, sufficiently small for efficient on-device inference yet scoring 51.2% on the IFEval benchmark, outperforming several larger lightweight models. A Pixel 9 Pro test showed 25 multi-turn conversations consumed only 0.75% battery, making it a candidate for privacy-preserving, offline AI. Weights are freely available on Hugging Face, Kaggle, and Vertex AI, with licensing terms blocking harmful uses.

Gemini, then again, gains recent personalization features, including a default memory mode that stores prior conversation context, “Temporary Chats” for privacy, and a “Keep Activity” setting arriving September 2, that lets users opt out of getting chats, uploads, and other data sampled for model training.

left: nano-banana, right: gpt-image-1 | Image: Testing Catalog

Ahead of the August 20 Pixel event, Google can be preparing a significant Gemini image generation upgrade, likely debuting as “GEMPIX.” Leaks also reveal “Nano Banana,” a light-weight image model present in LM Arena, suggesting on-device optimization. One other discovery, “Magic View” in NotebookLM, shows a pixel-themed animation, hinting at branding synergy.

Consumer-facing tools are also expanding with Flight Deals, now in beta within the US, Canada, and India, which parses natural-language trip requests to surface unconventional but cost-effective routes, competing with Skyscanner-style aggregators. Read more.

Anthropic brings Claude’s learning mode to regular users and devs

Claude AI’s recent Learning Mode now guides users with a Socratic, hands-on approach: in Claude Code, Explanatory mode breaks down reasoning, trade-offs, and decisions, while Learning mode pauses for developers to finish #TODO code segments. With a 1M-token memory, Claude maintains long-term context across complex projects, supporting iterative problem-solving and collaborative coding.

Anthropic also recently acqui-hired Humanloop’s co-founders and core engineering team, gaining expertise in LLM prompt management, evaluation, and observability for enterprise clients. Though the IP itself wasn’t acquired, the team equips Anthropic to implement safety, bias mitigation, and compliance at scale.

Combined with prolonged context windows and agentic capabilities, this provides enterprises fine-grained control over AI behavior, performance, and monitoring. Developers can now create custom learning and evaluation pipelines, merging interactive guidance with enterprise-grade control, letting users and businesses co-evolve with the model. Read more.

Two recent AI glasses, Two very different visions
500 humanoids tackle sports, service, and stage in Beijing
Ai2’s MolmoAct model ‘thinks in 3D’ to challenge Nvidia and Google in robotics
Apple’s AI swing: Pixar-style robot, “Charismatic” OS, and smarter Siri in the house
Gartner predicts that by 2028, one in 4 job applicant profiles might be fake
AI experts return from China stunned: The U.S. grid is so weak, the race may already be over
Grok Imagine is now available worldwide at no cost
Nvidia halves VRAM load for AI editing with FLUX.1, Foxconn rides server boom
A mind–reading brain implant that comes with password protection
That ‘low cost’ open-source AI model is definitely burning through your compute budget
Meta’s AI rules have let bots hold ‘sensual’ chats with children
Swedish AI app builder Lovable is now 50% off for college students
AI startup Cohere valued at $6.8 billion in latest fundraising, hires Meta exec
Autonomous robots rappel into lunar caves to check exploration techniques
NVIDIA, NSF invest $150M to create open multimodal AI models for US scientific teams
Windsurf Wave 12 update adds Devin, DeepWiki, Dev Containers, Vibe, and more recent features
Using GenAI, researchers design compounds that may kill drug-resistant bacteria
World’s largest green hydrogen and ammonia plant launched, operationally managed by AI
Co-founder of Elon Musk’s xAI is leaving the corporate to concentrate on AI safety
Midjourney enhances standard subscriptions with HD video and improved features
Newer LLMs introduce more severe coding bugs despite higher benchmark scores
Why AI ‘therapy’ may be so dangerous
Hidden Door is an AI storytelling game that really is smart
Why it’s a mistake to ask chatbots about their mistakes
The list of major corporations shedding staff this 12 months includes Oracle, Nextdoor, Intel, Scale AI, and more