I’m obsessed with robots and have been since I was a little boy watching Gigantor anime on my small tube TV. Back then, I dreamed of riding on the shoulders of my robot friend. I’ve since shed those boyish dreams in favor of a robot home companion along the lines of C-3PO, and part of me believes 2025 might be my year.
In 2024, we saw a lot of humanoid robot development from a growing number of companies. Market leader Boston Dynamics introduced the stunning Atlas 2, its most advanced and human-looking robot yet. Tesla trundled along with Optimus updates, Figure AI arrived with the remarkably bright Figure 01, and 1X Neo Beta’s uncomfortable human/robot interactions will forever haunt our dreams. The slow pace of humanoid robotic development I’ve witnessed over the last three decades has accelerated partly because some are pairing their robot efforts with advancements in AI. Figure 01, for instance, partnered with OpenAI for some voice-based responses.
When asked, as I sometimes am, when we will have Star Wars-like robots in the home, I usually avoid giving a direct answer because there are too many variables. Everything I know about robotic development tells me the wait will be long.
That, of course, is not the outcome I want. Robots remain at the topic of on my tech wishlist, so I thought that maybe I could find some moral and intellectual support from a few accomplices in the robotics technology race, namely, Google Gemini, OpenAI’s ChatGPT, Siri, and Anthropic’s Claude AI.
A simple robot question
I asked each of them the same simple question: “Will I have a humanoid robot in 2025”
Siri, even the Apple Intelligence upgraded one, was sadly useless. I kept trying to ask it this simple question, and each time, Siri thought I was in distress and told me, “If you think it could be serious, ask me to call Emergency Services or someone you trust.” I guess Siri assumed I was under robot attack.
I switched to ChatGPT 4o. As I mentioned, OpenAI is already deeply involved in the robot race, working with partners like Figure AI to enable smarter interactions with their existing humanoid robots. Perhaps I shouldn’t have been surprised by ChatGPT’s largely optimistic response.
ChatGPT told me it was “quite possible that we’ll have advanced humanoid robots. Technology is moving quickly, so you might see some impressive robots by then! Would you be interested in having one?”
“Affordable” is not in the 2024 or 2025 humanoid robotics vocabulary. I’m sorry, ChatGPT, but I don’t think you fully understand this market.
I was tempted to tell ChatGPT to “chill” as it asked me what type of robot I wanted. This is a question I found odd since I already told it I wanted a humanoid one.
ChatGPT’s belief in the rise of humanoid robots surprised me, so I challenged it and asked what made it possible for me to have one.
“Well,” it started in its most casual and, yes, human-like voice, “robotics technology has been progressing rapidly,” true, true. “Companies are developing sophisticated robotics with advanced AI,” there it is, “that can perform a variety of tasks. By 2025, it’s likely we’ll see even more practical and affordable humanoid robots.”
That last bit set off a red flag. “Affordable” is not in the 2024 or 2025 humanoid robotics vocabulary. I’m sorry, ChatGPT, but I don’t think you fully understand this market.
Things got more real when I switched to Claude AI and later, Gemini.
Anthropic’s AI threw some cold water on my dream, telling me, “In 2025, it’s unlikely that most people will have a fully functional humanoid robot at home,” adding that “truly versatile humanoid robots are still in early stages of development.”
Instead of Rosey the Robot, Claude AI tells me 2025 will feature:
- More advanced research and industrial robots
- Limited functionality robotic assistants
- More prototype humanoid robots in labs and tech demos
While I appreciated Claude AI’s more clear-headed view, I was also bummed out.
My dreams of a robot out in the backyard playing catch or cooking dinner while it regaled me with fantastic science fiction tales were fading.
Maybe Gemini would lift me up. After all, Google is the land of moonshot projects. Surely, they’d see the humanoid robot glass as at least half full.
Gemini, though, was even more taciturn than Claude AI. Here’s what it told me:
“While there are companies developing humanoid robots, such as Tesla’s Optimus and 1X’s Neo, it’s unlikely that they will be widely available to consumers in 2025. These robots are still in the early stages of development, and it may take several years before they are ready for mass production and commercialization.”
Image 1 of 4
Frustrated mainly because I know Gemini is right, I asked about “the main hold up,” and Gemini had quite a list. What surprised me is that AI sits at the very top of it. Gemini said that building artificial intelligence that can “understand and respond to complex environments [read unpredictable homes] and tasks in a human-like way is still a significant challenge.”
Other challenges include:
- Hardware that can handle all the different kinds of tasks we do in a home.
- Battery tech that can last long enough.
- Cost: It costs not only millions to develop and build these robots, but any humanoid robot you might try to buy in 2025 will cost tens of thousands of dollars.
It’s a harsh dose of reality from at least two of the world’s leading AIs and a lot of wishful thinking from one that should know better.
Despite all this, I look forward to an exciting 2025 in humanoid robotics. Sure, we won’t bring Optimus or Atlas home, but I guarantee we will see some eye-opening updates. That fun will probably start at CES 2025. At last year’s massive tech event, I danced with a humanoid robot. Who knows what they’ll be capable of this year.
@techradar
♬ original sound – TechRadar
