Martin Erlic, who runs an olive oil business in Croatia, used AI text-generating startup Jasper to write his dream sci-fi novel. Architect Nuno Fontarra turns to AI image software Midjourney to dream up whimsical buildings that inspire his real-life designs. And in moments of emotional turmoil, Serena, a physicist in Texas, asks the artificially generated life coach on Character.AI to help soothe her anxiety.

Silicon Valley futurists have long speculated—some with fear, others with giddiness—about what the world will look like when artificial intelligence infiltrates the mainstream. For years, their deepest hopes and fears failed to materialize. But in recent weeks, the tech industry declared in concert that the future had arrived. “With each of the previous waves, it died out. It wasn’t working. It wasn’t getting results,” said Ben-Zion Benkhin, founder of Wombo, a startup that uses an image generation model based on Stable Diffusion to manipulate photos. “This time is different. This time is really, really different.”

Generative AI technology—broadly defined as artificial intelligence that doesn’t just process preexisting data sets, but creates wholly original text, images, audio, videos and code—has never looked better, sounded crisper or been easier for the general population to access. As of last month, anyone can make an account on Dall-E 2 and create their own art, use Interior AI to completely redesign their living room or log on to Jasper and watch their computer churn out an essay in seconds.

And venture investors, with their billions, are hellbent on making the latest round of AI hype last: Jasper, a writing assistant, announced a $125 million fundraising led by Insight Partners at a $1.5 billion valuation; OpenAI, the powerhouse AI research company behind Dall-E 2, is in talks for a nearly $20 billion valuation, and just led a funding round for Descript, an audio- and video-editing AI startup, according to reporting by The Information; and Stability AI, the startup that uses open-source AI art-generator Stable Diffusion, just raised $101 million and threw a coming-out party in San Francisco that harkened back to startup booms of yore.

We spoke to dozens of users, engineers, founders and researchers who are deeply immersed in these tools. We then tried them out ourselves to assess their user friendliness and utility. What emerges is a shortlist of 14 services, products and tools that have leapt to the front of a very competitive—and increasingly very controversial—pack.