OneWave AI

Happy New Year! 🎉 As we welcome 2025, let’s take a moment to acknowledge the incredible progress AI has made in just a few short years. Since the release of ChatGPT in November 2022, OpenAI, Anthropic, Meta, and Google have been in an intelligence arms race. Here’s a condensed timeline of key AI milestones: 2022 • November: OpenAI launches ChatGPT, redefining conversational AI. 2023 • March: OpenAI introduces GPT-4 (multimodal AI), and Anthropic launches Claude (focused on safety). • May: Google upgrades Bard to PaLM 2, expanding global and multilingual capabilities. • July: Meta releases LLaMA 2, championing open-source AI. • October: OpenAI integrates DALL-E 3 into ChatGPT, merging text and image generation. • December: Google unveils Gemini 1.0 Ultra for advanced natural language understanding. 2024 • February: Google launches Gemini 1.5 beta with 1M token context length. • March: Anthropic debuts Claude 3 and its performance-leading Opus model. • May: OpenAI unveils GPT-4o, blending text, image, and sound processing. • July: Meta releases LLaMA 3.1, the largest open-source model yet. • September–December: OpenAI launches SearchGPT, ChatGPT Pro. Claude 3.5 sonnet is updated and Gemini 2.0 drops. Sora released to public. The speed of innovation is taking off! Need help deploying the right AI tools at your business for 2025?

Behind the scenes w/ John Lahr: Take a tour of our Mac-powered AI cluster running models locally. This is what real Local AI looks like ⚡️

In a recent podcast with Jason S. Weiss, Esq., Damon Lembi, CEO of Learnit, shared why OneWave AI stood out when they looked for an AI consulting partner: "I’ve explored working with several AI vendors and consultants, but OneWave AI speaks in layman’s terms. For someone like me, who’s not super technical, they help us focus on what matters: the outcomes and results. They’ve been phenomenal—the work they’ve done makes AI easy to understand, roll out, and adopt across the organization. And that adoption starts with senior leaders.” At OneWave AI, we believe AI success isn’t about complex jargon—it’s about: ✅ Clarity – Helping leaders and teams understand AI in simple, actionable terms. ✅ Outcomes – Focusing on measurable results that align with your business goals. ✅ Adoption – Building solutions that your teams embrace, ensuring real impact. Thank you, Damon Lembi and the Learnit team, for trusting us to make AI work for you. If you’re a leader looking to bring AI into your organization—and want a partner who simplifies the process and drives results—let’s connect. Watch 60 Second Clip Here: https://lnkd.in/eNBpEkxQ

This is pretty wild 👇 Fastest-growing software company to date? Rumors that StackBlitz has reached ca $20M in ARR two months after launching Bolt.new and now raising up to $100M at a +$1B valuation. Naturally challenging to measure actual ARR two months after launch with a monthly subscription-based pricing model. Wild nonetheless. AI-first software startups (leveraging e.g. Anthropic or OpenAI) that find product-market fit grow a lot faster than the previous generation of SaaS startups. Btw, for code gen tools like Bolt.new, Claude 3.5 Sonnet seemed to have been the big unlock: https://lnkd.in/dTh56tcJ To quote the founder of Bolt.new: "There’s an order of magnitude difference in the LLM’s required infrastructure to make it functional versus zero shot. Claude 3.5 Sonnet is the enabling technology that made this product possible, period." Interesting times! #startups #venturecapital #artificialintelligence

This is a major update🚨 This fit’s into and simplifies so many workflows for the average SMB or startup.

While OpenAI is doing their '12 Days of OpenAI', - which has so far primarily amounted to a new $200/month subscription - Meta AI just released Llama 3.3 70b, which outperforms GPT-4o on most major benchmarks and is ~13X CHEAPER for output tokens. In addition, it's already been implemented in Groq, giving you a ridiculous 1,600+ TPS. So to clarify: - open weight (awesome, love to see it) - 92.3% cheaper than GPT-4o and outperforms it across most benchmarks - ~10X faster inference than OpenAI via Groq - small enough to be run locally (within reason) The cost of intelligence is PLUMMETING to zero. 2025 is the year of the multi-agent framework, and low-latency, high-performance, open-weight models will reign supreme. I was initially skeptical of Meta when they went 'all in' on AI, but this is incredible work. Well done.

Fully autonomous or Human in the loop? Zapier’s CEO with a great clip on the value of “human-in-the-loop” AI systems. It’s a concept we deeply believe in at OneWave AI. AI is incredibly powerful at automating tasks and scaling operations, but its true potential is unlocked when paired with human intuition, creativity, and judgment. The balance between AI efficiency and human oversight ensures better results and greater trust. At OneWave AI, we design tools with this principle in mind: AI empowers, but humans elevate. Check out the video here: https://lnkd.in/e4fPCqwa #AI #HumanInTheLoop #Collaboration #OneWaveAI

AI-generated video just made a big leap. OpenAI’s Sora launched today. Here’s everything you need to know: - 1080p takes "a couple mins" (for now) - Explore feed is available to anyone, with or without a paid plan - Plus users get 50 generations, Pro users get 500 and unlimited slow queue generations - Remix, re-cut, blend, and loop existing videos - Releasing with “Storyboard” so you can "direct video with multiple actions across multiple sequences” - Edit your images into videos using Storyboard - Multiple resolution sizes are available, but “fewer generations at higher resolution” - You can “fine-edit” videos by setting strength of edits - Can generate up to 20 seconds of video Sora will be live today for most regions—excluding Europe/UK. Access here 👉 https://sora.com/

OpenAI o1 is now out of preview in ChatGPT. What’s changed since the preview? A faster, more powerful reasoning model that’s better at coding, math & writing. o1 now also supports image uploads, allowing it to apply reasoning to visuals for more detailed & useful responses. Plus and Team users will have access to OpenAI o1 today through the model selector, replacing o1-preview. Enterprise and Edu users will have access in one week.

⏱️ 5 minute tutorial: Learn how to build simple AI agents for content creation and digital design in just 5 minutes using our NVIDIA NIM microservices—powered by AI at Meta's Llama 3.1 405B 🦙 and stable diffusion models. 📹 https://lnkd.in/g5CNgQCe