It’s only open source if the training data is and it probably isn’t, is it?
I don’t know, though DeepSeek talk of theirs being “fully” open-source.
Part of the advantage of doing this (apart from helping bleed your rivals dry) is to get the benefit of others working on your model. So it makes sense to maximise openness and access.
I lived in Hong Kong for a few years. It has superlative public transport, and the (human) taxis were reasonably priced. However, as its so densely populated, I can only see cars getting so much traction. After a certain point the traffic jams are unavoidable.
I don’t know the specifics. What seems more relevant to me is that lots of automakers around the world are getting to Level 4 by various, mostly similar ways.
Once you have Level 4 you have a viable robotaxi business model. Even if you stick to geo-fenced areas, and mapped routes, that covers 80%+ of urban taxi journeys.
The same holds true for buses and public transit. I’m very interested to see how efforts like this Level 4 mini-shuttle bus in France progress.
When robotaxis & mass transit like these are common, how many people will still want private cars?
Base fares start as low as 4 yuan (around 55 cents), compared with 18 yuan (around $2.48) for a taxi driven by a human
China is already the global leader in 21st century energy - dominating renewables, batteries, and EVs. Now it’s poised to lead in robotic vehicles too.
Its robotaxis cost $30k, Waymo, who’s been in the robotaxi game longer costs $150k. Combine this with the fact Baidu can offer fares that are just 20% of a human driver in China and still make money and you can see how the global demand for such vehicles could be in the hundreds of millions. Tariffs in Europe and America may slow things there, but it won’t be the case for much of the rest of the world. Cheap Chinese robotaxis, with fares a fraction of today’s human-driven journeys, will be ubiquitous all over the planet in the 2030s.
We rarely hear of the Chinese space program in western media, but it keeps doing interesting things. A recent launch tested an inflatable module for their space station. That was an idea that once seemed promising for the ISS, via Bigelow Aerospace, but never seemed to go anywhere.
This cargo mini-shuttle concept isn’t new either. Thirty years ago an ESA version called Hermes got to the advanced planning stage before being scrapped. Some people have doubts that space planes, even launched with reusable rockets, are all that efficient, so it will be interesting to see how this fares.
I wonder what the monthly totals will be next year. When will it hit the 1 million mark, then 10 million?
The problem for all the investor funded AIs, is that data centers are huge costs. They’re burning through billions of dollars every month. That makes sense if one of two of them emerge as dominant players who own most of the market share for future AI businesses.
If they all keep under-cutting each other by using open-source. It’s more likely companies like OpenAI will crash and burn first.
Also, to me this means future self-driving transit will be commoditized.
If robotaxis can be as cheap as $10K, then their main cost per passenger journey will be the electricity to run them.
A lot of these open sourced AIs have been released that way to undercut competitors. Many people still think AI will consolidate into the hands only the biggest of Big Tech players, but it doesn’t seem to be happening yet.
They are also capable of making decent cars they sell for as cheap as $10-15,000.
Also, current LLMs are great at modelling best practices.
Most disease diagnosis, even rare diseases, follows predictable paths. Human doctors would have to have superhuman memories to do as well.
What’s more exciting to me is that this knowledge is now free. Free as in beer.
People talk of UBI, but what about universal services that cost nothing?
It’s a no from me. I suspect as the US gets more deregulated for AI, it will be more no’s from people around the world.
Now that scaling is hitting a wall, it will be interesting to see what methods like this continue AI’s progress.
The company’s website says they are using generative AI too.
They are using it to generate virtual environments to train the self-driving AI. Waymo is using generative AI to do the same.
This will be a great way to channel vast sums of money from the American taxpayer to rich elites, for which the taxpayer will see little or nothing in return. Something the US public are about to see a lot more of.
While people usually focus on carbon neutrality, I often think decentralization is renewables’ most underappreciated aspect. Everything it touches can happen at the home and community level. The Haber-Bosch process is the epitome of the 20th century large scale heavy industry model. Now here is a solution replacing it at the level of individual farms.
I suspect much of robotics will be decentralized too, and with that, they may decentralize automated manufacturing. In a few decades, it may seem quaint that people shipped so many things halfway around the world.
Now that the new US administration is about to gut AI regulations, this idea gets even worse.
He said it again a few days ago on a Reddit AMA.
Perhaps the most interesting comment from Altman was about the future of AGI - artificial general intelligence. Seen by many as the ‘real’ AI, this is an artificial intelligence model that could rival or even exceed human intelligence. Altman has previously declared that we could have AGI within “a few thousand days”.
When asked by a Reddit user whether AGI is achievable with known hardware or it will take something entirely different, Altman replied: “We believe it is achievable with current hardware.”
Why are they making it so needlessly complicated? They can just use existing highways and vehicles with Level 4 self-driving. They don’t need new separate roads.
That said, this points to the future. Even if true Level 5 self-driving is several years off, there is plenty Level 4 can do now. That includes all cargo driving on highways. I doubt most trucker jobs have long to go. Some will say they are needed for last-mile delivery. Some companies are soon going to figure out a profitable system for having human drivers locally for that, but self-driving vehicles for the long stints on highways.
/cdn.vox-cdn.com/uploads/chorus_asset/file/25749641/2165011649.jpg)
There’s a few ways they say it may help, this one seems the main one.
We foresee a future in which LLMs serve as forward-looking generative models of the scientific literature. LLMs can be part of larger systems that assist researchers in determining the best experiment to conduct next. One key step towards achieving this vision is demonstrating that LLMs can identify likely results. For this reason, BrainBench involved a binary choice between two possible results. LLMs excelled at this task, which brings us closer to systems that are practically useful. In the future, rather than simply selecting the most likely result for a study, LLMs can generate a set of possible results and judge how likely each is. Scientists may interactively use these future systems to guide the design of their experiments.