Welcome to ZDNET’s Innovation Index, which identifies the most innovative developments in tech from the past week and ranks the top four, based on votes from our panel of editors and experts. Our mission is to help you identify the trends that will have the biggest impact on the future.
We’re joined by three new panelists this week: CNET’s Connie Guglielmo and analysts Carolina Milanesi and Avi Greengart.
In first place, after last week’s predictions around agentic systems and investment in AI, the trend continues: AI will take on more work over the next decade until, yes, it replaces humans in certain jobs. ZDNET contributor Vala Afshar lays out his six levels of autonomous work as a framework for what to expect in the next 10 years based on industry. Initially, AI will continue to augment human labor, or what we see today: time saved, opportunities opened. Not linearly, but eventually — in roughly five years, according to Afshar — AI will be able to perform some roles currently overseen by humans. Afshar lays out his recommendations for businesses and HR teams navigating these changes; read the full piece for tips on how to prepare.
Coming in at #2 are AI voice generators and the countless parts of our lives they’ve come to touch. As ZDNET’s resident AI deep-diver David Gewirtz notes, synthesizing voices isn’t new — attempts date back to 1791 — but the scale and precision AI makes possible are. The implications of this are endless. Whether used to help ALS patients speak or scam you out of some money, AI-generated voices are everywhere and becoming harder to distinguish from human ones (which panelist Connie Guglielmo is extra watchful of as we creep towards a misinformation-riddled election). While that’s an eery reality, it has its pros and cons — especially, as Gewirtz found, when calling customer service.
Also: AI chatbots too English-language centric, Stanford study finds
In third place is the launch of three new experimental Gemini 1.5 models, which Google released this week to much fanfare (and some grumbling). The DeepMind team lauded 1.5’s talents in an accompanying technical report, noting the model family’s “near-perfect recall” of almost 99%, “unprecedented” context windows, and its ability to process nearly five days of audio. While internal research is always worth outside vetting, the new models made waves in the Chatbot Arena within hours of the launch, beating out competitors for spots #2 and #6 overall.
Closing out the week is Claude, Anthropic’s steady — demure, even — chatbot, which got an impressive update. The assistant’s new Artifacts feature, now generally available, formats its responses to your queries in real time, making it easier to create and implement code snippets, graphics, dashboards, and more. ZDNET contributor Lance Whitney called it the “coolest feature” he’s seen in generative AI so far. Keeping with the theme of recent subtle yet effective AI features, this consumer-friendly update makes getting the most out of Claude more seamless for the everyday user, which could mean increased adoption for additional tasks.
Artificial Intelligence
