Demystifying Private GenAI solutions

Investments in the AI industry reached astronomical highs with the $300bn deal between OpenAI and Oracle. Competition reached the governmental level with the announcements of investments: US $500 bn, EU over $200 bn and China aims to reach $98 billion by the end of the year. On the technology side, GPT-5 was recently released. Unlike prior versions, it did not represent a paradigm shift but rather an incremental update with improved test results. This development deviates from the scaling-performance trend and echoes Yann LeCun's (MetaAI) statement, suggesting that artificial general intelligence (AGI) will not be achieved merely by scaling large language models (LLMs). Furthermore, Apple's latest research highlighted LLMs' limitations in mathematical reasoning. Another study by University College London acknowledges the LLM’s "scaling wall" issue, leading to a significant increase in computational costs for error correction. Therefore, the question remains: can LLMs innovate and acquire new skills as a "PhD in a Pocket"? Let’s take a moment to explore how current AI technologies can benefit project managers. While off-the-shelf generative AI solutions, which we discussed in the previous report, are quietly making their way into our office suites and smartphones, today, we will focus on private generative AI solutions. These solutions include not only data preparation and training, but also hosting infrastructure and development or customisation of the GenAI model. Gartner are sceptical about whether this way would be selected by the majority, and shortlists those who can choose it: ·Corporates; ·Software Product Development companies, including Startups; ·Niche businesses that haven’t found the right off-the-shelf solution and are willing to develop their own solution. Private GenAI solutions require strong expertise in software development, data, testing, hosting infrastructure, implementation, training and, as always, support. Benefits: Private GenAI enables exploration, innovation, and modification of nearly any use case in project management, resulting in solutions that can surpass off-the-shelf options. As we fine-tune the model itself, it also allows combining multiple models. And offers the next-level security with full control over its data infrastructure, data flows, and models. Trade-offs: Across different sources, the failure rate of GenAI PoCs in business is 80-90%, and it can reach a shocking 95% for solo implementations, according to recent research by MIT. So companies should be very selective and carefully evaluate the outcomes and future-proofing of their PoC’s use cases. Responsibility for ensuring compliance with the EU AI Act and GDPR regulations. If the organisation seeks a solution integrated with internet search or one that would work with multiple output modalities, it is essential to consider RAG and Agentic-AI solutions closely. By Denis Makarov, who is the IT Solutions Program Manager at Sanbra Group Ltd

Google’s New Gemini 3 AI Crushed OpenAI and Anthropic in a Benchmark Test for Business Operations

Google has released Gemini 3, the latest in its line of advanced AI models. As most AI companies do when announcing a new flagship model, Google boasted that Gemini 3 is its most intelligent model yet, and tops several benchmarks, including one that judges an AI’s ability to run a business. Google has also released a new application to supplement Gemini 3’s coding power. After months of teasing, Google CEO Sundar Pichai finally announced Gemini 3 in a blog post, saying that it enables anyone to “bring any idea to life.” The model is now integrated throughout much of Google’s ecosystem, including its search engine’s AI Mode, Google AI Studio, and the Gemini App. Pichai said that Gemini 3 is “much better at figuring out the context and intent behind your request, so you get what you need with less prompting.” Gemini 3 will be a family of models that vary in size and price. For now, the only model available is Gemini 3 Pro, which is the largest and most expensive version. Over time, smaller and cheaper versions of the model will be released. Gemini 3 Pro also includes a “Deep Think” mode, which has become standard across AI platforms. By activating this mode, Gemini can think even longer and harder about how to solve complex problems. Demis Hassabis, CEO of Google DeepMind, wrote that Gemini 3 is “the best model in the world for multimodal understanding and our most powerful agentic and vibe coding model yet, delivering richer visualizations and deeper interactivity.” By multimodal, he’s referring to the capability of AI models to process and generate content across a variety of mediums, including text, images, and video. Vibe coding refers to the practice of directing AI agents to write and execute code on your behalf, and has been a major AI topic in 2025. In its blog post, Google also claimed that Gemini 3 Pro is significantly less sycophantic than other AI models. “Its responses are smart, concise and direct, trading cliché and flattery for genuine insight,” the company wrote, “telling you what you need to hear, not just what you want to hear.” According to Google’s own testing, Gemini 3 Pro tops several widely-used AI benchmarks, including MMMU, which gauges multimodal understanding, and Terminal-Bench, which judges a model’s ability to code within a computer terminal. One notable leaderboard that Gemini 3 Pro topped was Vending-Bench 2, a benchmark that measures an AI model’s ability to run a business (in this case a vending machine) over a long period of time. After a full simulated year of operation, Gemini 3’s bank account balance was $5,478.16, much higher than second place finisher Claude Sonnet 4.5, which ended the virtual year with $3,838.74. Google clearly has high hopes for Gemini 3 in the coding domain. Along with the new model, the company has released Google Antigravity, a new agentic development platform that will likely compete with fast-growing startup Cursor, which sells its own AI-powered integrated development environment (IDE). Google Antigravity gives AI agents access to a code editor, terminal, and browser. In addition to Gemini 3, Google Antigravity users will also be able to select Anthropic’s Claude models and OpenAI’s open-weights model. Google says that Antigravity also comes “tightly coupled” with Nano Banana, the company’s popular image-editing model. For nontechnical founders who might be intimated by the technical details of Antigravity but want to try their hand at AI coding, Google has brought Gemini 3 Pro to Google AI Studio, a web-based application designed specifically for those without coding experience. In a blog post, Google AI Studio product lead Logan Kilpatrick wrote that Gemini 3 Pro “can translate a high-level idea into a fully interactive app with a single prompt. It handles the heavy lifting of multi-step planning and coding details delivering richer visuals and deeper interactivity, allowing you to focus on the creative vision.” Gemini 3 Pro is currently available for enterprise use for members of Google’s Gemini Enterprise platform. Google says that several businesses are already using Gemini 3 Pro, including Box, Cursor, Harvey, Replit, Thomson Reuters, and Shopify. Gemini 3 Pro costs $2 per million tokens on input prompts that are smaller than 200,000 tokens, and $12 for per million tokens generated. Tokens are units of data that are processed and generated by AI models. BY BEN SHERRY @BENLUCASSHERRY

10 AI Tools Marketers Are Using Right Now

Who doesn’t want to be more efficient? That’s why millions of Americans are turning to AI at the office. Over the past year, the share of U.S. workers using AI tools as part of their job doubled to 40 percent, according to a Gallup poll published in June. That number gets even higher for the marketing industry. By one measure, 61 percent of creative and marketing professionals use AI for their work, including analytics, content creation, strategy, and planning. Another poll found that 76 percent of marketers employ AI tools. For those marketers searching for the most effective AI tools to incorporate into their daily work flow, Inc. surveyed marketing-focused founders and their chief marketing officers. Here are the models and platforms they cannot live without. 1. ChatGPT Unsurprisingly, one of the most frequently cited tools was ChatGPT. Founders and chief marketing officers rely on the LLM for daily tasks, including brainstorming, market research, data analysis, strategy development, and content creation. One chief marketing officer trained a custom GPT to become chief of staff. Another says they use ChatGPT to formulate step-by-step guides when learning how to use other AI tools. With a team of fewer than 10 people, Sophie Mann, chief marketing officer of Furnished Finder, an online marketplace for furnished rental properties, uses ChatGPT as both her executive assistant and copywriter. Mann taps the LLM to help her draft board updates, structure meeting agendas, write performance reviews, and negotiate partner contracts. Without a full-time copywriter on staff, Mann also built a custom GPT, which is trained on Furnished Finder’s brand voice and customer personas, to write content. “It’s now the starting point for nearly every marketing asset. From email campaigns, social captions, blog posts, and paid ads, to event collateral and even voice-over scripts for our phone lines. You name it. We’re likely starting our drafts in Chat,” says Mann. “These tools help us scale our output without adding headcount. Reading through 1,000+ customer survey responses, for example, used to take hours. Now, I can have AI summarize key themes and insights in under a minute.” 2. Descript For creators and founders building in public through video, whether it’s TikTok or podcasts, Descript has become a go-to tool to make editing easier and faster. The AI-powered video editing platform, which landed on Fast Company’s list of Most Innovative Companies this year, creates transcripts from raw video and lets users edit through a text document by deleting words, phrases, or entire chunks. Last year, Descript added new AI capabilities, which automatically remove filler words, repeated words, bad takes, and background noise. Overall, the company claims its tool enables users to make “130 percent more videos in 27 percent less time” and “first-time users were 25 percent more likely to complete their project,” Fast Company reported earlier this year. 3. OpusClip Many of the founders who use Descript also use OpusClip, an AI-powered video editing software that helps users cut down longer videos, such as hour-long podcast interviews, into a series of shorter clips for social media. Within two years of launching, the company has scaled to more than 12 million users and become a favorite of social media managers. Shana Ayabe, founder of the marketing company Grace Digital Media and co-host of the podcast The Exit Interview, calls the tool a game changer. Opus Clip “allows us to quickly edit, reformat from landscape to vertical, and add dynamic captions while maintaining high production quality,” says Ayabe, who uses the paid version so her team can collaborate in real-time on the platform, “hearting” and commenting on different clips. “The built-in virality scoring system helps us understand why a clip is likely to perform well, so we can strategically schedule posts around traffic patterns and trends.” 4. Perplexity ChatGPT is not the only LLM that marketers use. In fact, most of the founders and CMOs who spoke with Inc. use multiple different models, depending on the task. Perplexity was the tool of choice when it came to research, including analyzing documents and data sets. Denise Aguilar, a global marketing strategist and founder of her eponymous Seattle-based company, Denise Aguilar Consulting, used the paid version of Perplexity and says the LLM has streamlined her workflow, allowing her to take on more ambitious projects for her clients. “The upgraded features, such as advanced file handling, faster processing, and priority support, have enabled me to work with large sets of PDFs and rapidly search, organize, and synthesize information,” says Aguilar, who has worked with companies, including Microsoft, Amazon, General Motors, and Vogue. “Investing in the Pro version has definitely raised my efficiency, especially when refining communications strategies, building personas, and editing across multiple marketing campaigns.” 5. Midjourney Midjourney, an AI-powered image generation tool, is being sued by a collection of visual artists and the major Hollywood studio Warner Bros. for copyright infringement, but marketers, especially those who work in the creative side of advertising, still say the tool is helpful for developing concept art and mock ups. To avoid any legal issues, be careful to restrict any images to internal and exploratory use only. 6. Lovable Marketers have joined the vibe-coding trend and started developing their own software by telling AI tools what they want to create, rather than writing code themselves. Many founders and CMOs prefer to use Lovable. The platform has become so popular that it became a unicorn within eight months of launching and has attracted nearly eight million users. Inc. AI reporter Ben Sherry used the free version to create an entire website in an hour. Marketers tend to opt for the paid version and say the tool is especially helpful for creating prototypes of client websites and apps. Maria Pergolino, chief marketing officer of SPS Commerce, a Minneapolis-based software company that helps retail partners optimize supply chain operations, says Lovable has been “transformative” for her work flow. “No longer do you need to awkwardly describe your vision for an app, ad, slide, or campaign. Instead you can describe your vision to an LLM and pop the directions into Lovable to bring your ideas to life,” says Pergolino. “This saves me hours every week.” 7. Claude If Perplexity has become the researcher and ChatGPT has become the catch-all for marketers, Claude has become the go-to LLM for writing. Founders and CMOs say the model excels as a place for brainstorming, storytelling, and testing out ideas or phrases. Patrick Finan, the co-founder and CEO of Block Club, a branding, strategy, and content agency for B2B technology companies based in Brooklyn, says Claude is the main LLM that he and his team use. “It’s fully integrated into Slack, Gmail, Google Workspace, Google Calendar,” he says. 8. Gamma For help making presentations, marketers have turned to Gamma as the PowerPoint of the AI era. The AI-powered platform takes text, such as documents or outlines, and transforms them into a slide presentation with one prompt. Using this same method, Gamma also lets users create polished-looking PDF documents, social media assets, and websites. Within two years of launching the San Francisco-based startup has attracted 50 million users, Fast Company reported earlier this year. Founders who spoke with Inc. recommended the paid version. 9. AirOps “Marketers are losing their minds” trying to optimize their existing SEO strategy for the new era of AI search, Andy Crestodina, co-founder and chief marketing officer of Orbit Media Studios, a Chicago-based digital agency that focuses on web development and website optimization, told Inc. recently. AirOps has helped streamline that process, founders and CMOs say. The company, which secured a $40 million Series B fundraising round earlier this month, calls itself the first content engineering platform for AI search. Marketers say the platform makes their AI optimization strategy more efficient. Leah Taylor, who runs communications for the AI sales platform Apollo, says Apollo has embedded the AirOps AI infrastructure into its core marketing operations to automate performance reporting and identify new revenue opportunities. “AirOps ingests our first-party data across notes, OKRs, experiment logs, and Slack, then uses enterprise LLMs to analyze and publish insights,” says Taylor. Since WebFlow, a no-code website experience platform, started using AirOps, the company increased its visibility on AI search with more than 330 new citations and a 24 percent uptick in SEO impressions, says chief marketing officer Dave Steer. AirOps has also increased revenue, turbocharging AI-attributed signups jumping from the low single digits to nearly 10 percent. 10. Gemini Marketers who are incorporating LLMs into their daily workflow also name-checked Gemini. While Doug Straton, chief marketing officer at Bazaarvoice, an Austin-based software platform that helps brands harness user-generated content, ratings, and reviews, calls ChatGPT the “easiest and most fun” LLM to use, he usually turns to Gemini instead for its repeatability and reliability. “I find Gemini, while harder to brief, creates more uniform, consistent results. It’s my company’s default,” says Straton, who uses the paid version. “It seems less eager to please you with a result you want to see, versus what you need to see.” BY ALI DONALDSON @ALICDONALDSON

What’s Next for AI? Andreessen Horowitz Founders Share Their Thoughts

Stocks of companies tied to artificial intelligence have been hitting stratospheric levels for over a year now, thrilling investors, but also causing concerns about a potential AI bubble. As startups close breathtaking funding rounds, like the $40 billion OpenAI collected in March of this year, fears of an AI bubble are growing — and some say a burst could be even bigger than the dot-com bubble of the late 1990s. The bubble theory is hotly debated. Some within the industry say they agree that the investment landscape is bloated, including OpenAI co-founded Sam Altman. Other experts, like Goldman Sachs, however, say we’re not in one (yet) — and Fed chair Jerome Powell has been skeptical of the bubble calls. As that debate rages, investors continue to fund AI startups. Few investors are in as deep as Marc Andreessen and Ben Horowitz. Their venture firm, Andreessen Horowitz (commonly called a16z), has sunk billions into the AI space. In April, it was reported the company was in early talks to raise a massive $20 billion AI-focused fund. The two investors recently came together at a16z’s Runtime conferences to talk about where AI can go beyond chatbots. Neither was willing to make any specific predictions about AI’s forthcoming capabilities, saying it’s too early to even imagine that. Andreessen likened AI to the personal computer in 1975, noting there was no way at that time to imagine what PCs would be capable of today. However, he expects similar levels of advancement — from a stronger starting point. AI, he said, is already approaching levels of human creativity — and while Andreessen would love to see humans continue to have superiority in that area, he thinks it’s unlikely. Tools like OpenAI’s Sora 2 video, for instance, are already capable of creating realistic scenes, animations, and special effects — and the introduction of AI actress Tilly Norwood has caused an outcry and prompted debate in Hollywood. “I wanna like hold out hope that there is still something special about human creativity,” he said. “And I certainly believe that, and I very much want to believe that. But, I don’t know. When I use these things, I’m like, wow, they seem to be awfully smart and awfully creative. So I’m pretty convinced that they’re gonna clear the bar.” Horowitz agreed, saying that while AI might not currently create at the same level as human artists, whether painters or hip-hop performers, that’s largely due to how little it has learned so far. It’s just a matter of time before it has an equal or superior level of talent. And some artists are already looking to use AI to collaborate, he said. “With the current state of the technology, kind of the pre-training doesn’t have quite the right data to get to what you really wanna see, but, you know, it’s pretty good,” he said. “Hip-hop guys are interested because it’s almost like a replay of what they did — they took other music and built new music out of it. AI is a fantastic creative tool. It way opens up the palette.” While AI can devour as many data sets as programmers throw at it, that doesn’t give the technology situational awareness. It is, in essence, book smarts versus street smarts. But the robotics field is expanding quickly. Elon Musk and Tesla are working on humanoid robots and Robotics company 1X has already started to take preorders for a $20,000 humanoid robot that will ‘live’ and work around your home. Once that technology and AI are blended, Andreessen said, AI will see a significant jump in actionable intelligence. “When we put AI in physical objects that move around the world, you’re gonna be able to get closer to having that integrated intellectual, physical experience,” he said. “Robots that are gonna be able to gather a lot more real-world data. And so, maybe you can start to actually think about synthesizing a more advanced model of cognition.” While there are plenty of experts who warn the AI market could be in a bubble right now, including OpenAI CEO and co-founder Sam Altman, Horowitz dismisses the idea, saying bubbles occur when supply outstrips demand — and that’s not the case with AI. “We don’t have a demand problem right now,” he said. “The idea that we’re going to have a demand problem five years from now, to me, seems quite absurd. Could there be weird bottlenecks that appear, like we don’t have enough cooling or something like that? Maybe. But, right now, if you look at demand and supply and what’s going on and multiples against growth, it doesn’t look like a bubble at all to me.” BY CHRIS MORRIS @MORRISATLARGE