Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI. learn more
Why read AI predictions from someone who isn’t Sam Altman or an AI Ph.D.? Well, early stage VCs spend their days collaborating with people far smarter than us who use AI at the cutting edge of what’s possible. I’m spending time. This allows you to know what will happen in the future.
In over 30 years of experience in technology, I have never seen advances in technology, especially in and around fundamental models, have such a huge impact in so many ways and so quickly. This has reshaped my expectations of how teams are built and how companies deliver value.
With that in mind, here are my predictions for 2025.
Attempt to contain AI
I love Mustafa Suleiman Consequences of the digital species Because AI frames the containment problem well. Much of the current AI built into products suggests courses of action. But AI is increasingly taking actions on our behalf. we have to frame ours please don’t I want you to do that.
Expect cage rattling to be loud in this area. Using California’s failed SB1047 bill as a bellwether, Elon Musk’s pro-regulatory stance and Trump’s anti-regulatory stance could offset penalties. Look for something like Anthropic’s Constitution model on steroids, as the industry attempts to “self-regulate” the safety of its underlying models to avoid government regulation. Similarly, we expect to see the rise of AI security startups that support cyber containment.
Basic model as brand DNA
Some of you may remember when “Intel Inside” was a big campaign. The foundational model brings brand power to the products built on top of it. This goes beyond tone of voice, and according to Claude, today’s tone of voice is:
- GPT-4: A good and careful consultant.
- Claude: A prudent and ethical advisor.
- Llama: A rough and flexible generalist.
Tone is adjustable. How understanding and recall are applied to models differ. The model will inevitably shape the “feel” of the product.
AI for the Nobel Prize
Although not perfect, it greatly improves our ability to repeat scientific discoveries. Today, AI collaborates with humans to enable breakthrough advances. i agree Vinod Khosla About this: AI scientists (super-expert AI agents) will regularly outperform human scientists in discovering insights.
Because AI scientists don’t need breaks. They have infinite patience for repetition. They have perfect memory for data. And they can work 24/7. Many startups are helping companies leverage this relentless work ethic, and let’s take a look at the first glimpses of this relentless work ethic in the advancement of medical discovery and healthcare.
Audio interface redefines “hands-free”
If Alexa, Siri, and Google Assistant were all v1.0 audio interfaces, audio would be accompanied by the additional context of visuals and text, creating entirely new product experiences that recalibrate our standards.
Specialized audio-only foundation models are competing with larger foundation models to push the possibilities here. We’ve already seen how Meta’s smart glasses have come a long way and how far they’ve come from Google Glass. “Hands-free” operation takes on a whole new meaning and expectation.
Overcoming the trust tipping point with Gen AI
I have a good sense of direction. Only a few times I got stuck in traffic that I didn’t know about, and the map app did, but eventually I surrendered, believing that the map app knew the route better than I did.
The same thing happens with generative AI output. To this day, experts continue to produce excellent results. This gap will narrow and we will succumb to the output of more generations of AI. That’s because we assume they are at least as good as, and in some cases better than, experts. They can get credit the same way we cite them. Case in point: Check out the branding section of this article.
we pay a premium for handmade
Non-digital products with evidence of human creation will become more valuable as the barrier to creation becomes lower and Dall-E and many other tools can create new designs in seconds. Think not only about original photography and visual design, but also about ceramic mugs made by artists rather than die-cut by robots in a factory.
My adult children marvel at VCRs and love vinyl records, but this is beyond vintage cool. In a highly digitized world, analog and physical meaning emerges in surprising ways, especially for Gen Z and Alpha.
“A human was found here.”
Many of my friends like Waymo because they don’t have to talk to anyone. However, when the time comes, it is important to have the ability to interact with humans who can make decisions that are not based on black box logic. The difference is whether an AI or human-driven approach is more valuable.
Expect backlash in areas such as customer service, where the rapid shift to AI has left many frustrated. Human-involved systems are considered preferable to fully autonomous systems. Businesses (and consumers) will promote and value that difference.
What a year…
I leave 2024 humbled and 2025 expected to be a chaotic year. But I’m an optimist and I believe wholeheartedly in the fact that we’ve built great companies and continue to surround ourselves with really smart people who are trying to make things better. I pray that the coming year will bring you humility, joy, and lots of blessings.
Martina Lauchengco is the next partner. Costanoa Ventures.
data decision maker
Welcome to the VentureBeat community!
DataDecisionMakers is a place where experts, including technologists who work with data, can share data-related insights and innovations.
If you want to read about cutting-edge ideas, updates, best practices, and the future of data and data technology, join DataDecisionMakers.
Why not consider contributing your own articles?
Read more about DataDecisionMakers