San Francisco Bay Area
Joined May 2020
- Tweets1.7K
- Following696
- Followers254
- Likes286
Expected. The only advantage of apple devices is its unified memory architecture and I was hoping strides in the GPU architecture. To be fair they did improve, but they’re 3-5 years behind schedule and NVIDIA and the rest of the industry won’t wait for them.
Hinton got a nobel prize, invented/popularized back-propagation and AlexNet and basically helped bootstrap this age of AI. Despite of that have been completely wrong about every prediction he said about the future of AI and its implications to future of humanity. A 27y old senior researcher at Anthropic is levels below him and capable of make foolish extrapolations as well
Ganhar o prêmio nobel não te faz um especialista em extrapolações sobre o fim da humanidade. Na linguagem popular o Hinton está “se passando” há mais de uma década
And I think the timing of your post and the people quoting it looks extremely coordinated and suspicious.
I don’t think you’re an optimist. The attacker X Defenders game is unbalanced right now, but it won’t be the case in the long run. Hinton also made outrageously wrong predictions due to extremely biased linear extrapolations. I think doomers are taking us nowhere. We need more realistic, but hopeful people leading the area.
Raphael Amorim retweeted
🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient.
🔹 Introducing the smallest model in our new architecture family, with native visual understanding.
🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models.
1/6
Raphael Amorim retweeted
GLM 5.3 Flash is 0x Alpha for those that are curious
Tomorrow it's going to be a great day for Local AI with Qwen and GLM releasing new model weights.