Tetsuya

Co-founder ($15M+ ARR) ⬅ AI engineer @SAP
I'm interested the individual variability in neural activities.

[Tetsuya @_HaneAI]

Office of the lab I’m researching.

We’ve researched and developed Neuromodulation and Biosignal computations. #BCI

[佐藤航陽（さとうかつあき） @ka2aki86]

国や業界が新しいものを生み出すのを諦めても、自分がやることは変わらない。新しい世界を望む人を探し出して、未来の可能性を追求するだけ。他人の閉塞感を受け取るかは自分が選べる。

[Tetsuya @_HaneAI]

Wanna try a simple function that predicts the user's next action based on retinal movements, which is a feature of #VisionPro. Predicting motor control is one I've kept eyes on in prior research.

Waiting for the right time and putting it into products with a reasonable quality.

[Sterling Crispin 🕊️ @sterlingcrispin]

One aspect of #VisionPro that I think people haven't fully been able to appreciate is the advanced Spatial Audio system. It's insane, and the whole device is really the most advanced consumer electronics device ever created.

The audio system creates the feeling that sounds are coming from the environment around you. The system measures the geometry of your ears and head with an iPhone depth camera to create a personal profile. It also analyzes your room's acoustic properties, including the physical materials. And when a sound is played it uses audio raytracing.

What's all that mean? Imagine you're in nature and you hear a bird chirp to the left of you. The sound waves the bird makes travel in all directions, bouncing off of trees, the grass, your body, and finally into your ears. Some of those objects dampen the sound, some reflect the sound, others bend the sound. And the sound takes a slightly different amount of time to reach each of your ears. Then the sound echos off of the unique shape of your ears, which changes its pitch, and all of this lets your brain know what direction the bird chirp came from, and also what else is in your environment.

So again, the advanced spatial audio system of Vision Pro knows the geometry of your ears if you've calibrated that with an iPhone, it analyzes the acoustic properties of your room including the physical materials, and simulates the propagation of sound using audio ray tracing. I don't think theres any other system on earth capable of this. When you're in a face time call, it sounds as if people are speaking right from where they are. That sounds obvious and like a simple feature, but this is impossible to fully appreciate without experiencing it.

George Lucas is famous for saying sound is more than half the experience when watching a movie. And there's really no way of fully communicating the spatial audio system without hearing it for yourself. Imagine putting on a high end pair of headphones to Vision Pro and watching a 3D movie with a screen that feels 100 feet wide.

[Tetsuya @_HaneAI]

OpenAI CEO Sam Altman said at today's hearing with the U.S. government that he does not own OpenAI stock.

[Tetsuro Miyatake @tmiyatake1]

今日のアメリカ政府との公聴会でOpenAI CEOのSam AltmanさんがOpenAI株を持っていないことを発言。

[Tetsuya @_HaneAI]

One paper said the more computer resources, data and epoch etc. are increased, the more key indicators like acc/val improve. After building it, I might understand it sensibly.

So excited to making some improvements around Encoder next.

[Tetsuya @_HaneAI]

Here's my boss's google scholar. He is an authority in this area, very nice and smart as well. https://t.co/OYZ3emVg7f…

Our lab have been doing extremely practical and challenging research for our future.

[Tetsuya @_HaneAI]

Office of the lab I’m researching.

We’ve researched and developed Neuromodulation and Biosignal computations. #BCI

[Tetsuya @_HaneAI]

Apparently you can watch the moment Agility Robotics, Inc.'s bipedal robot dies of exhaustion... This is going to make for more kinds of insurance for robots🫥

[ヴィズマーラ恵子🇮🇹 @vismoglie]

米国の会社 Agility Robotics によって作成された二足歩行ロボット、20時間以上、1日中働いて過労死。

[Tetsuya @_HaneAI]

Genome analysis on Beethoven's hair revealed that Beethoven had a genetic predisposition to liver disease and a hepatitis B infection before his death. The team of researchers, led by Cambridge University, analysed the his genome from five locks of hair. https://t.co/FoIIbFiz6K

[Tetsuya @_HaneAI]

It seems that ChatGPT is used as a function in the paper as y = ChatGPT(x).

It's too early to say, but machine learning is becoming more and more a huge system that's beyond human comprehension.

[ketsumedo_yarou @ketsumedo_yarou]

論文中でy = ChatGPT(x)のようにChatGPTが関数として使われているのを見てしまい技術進展が早すぎる

[Tetsuya @_HaneAI]

Worth a look for anyone interested.
FDA already announced their guideline for Invasive Brain-Computer Interfaces!!

It includes guidance on the biocompatibility of the materials used and electrodes etc.

[U.S. FDA @FDA]

Today, FDA issued final guidance on implanted brain-computer interface devices for patients with paralysis or amputation. This guidance is intended to facilitate the development of safe and effective products employing this emerging technology. https://t.co/F5nzFs4kKk

[Arib 🇺🇸🇵🇰 @aribk24]

I controlled a tesla with my brain 🧠

here is a demo I did in front of a 100 people

[Tetsuya @_HaneAI]

The time has come for artificial intelligence to be able to judge the output produced by artificial intelligence like ChatGPT.

Stanford introduces DetectGPT to help educators fight back against ChatGPT generated papers #stanford #chatgpt #ai #openai: https://t.co/GYChxYkMes

[Tetsuya @_HaneAI]

Therefore, they had Kenyans do the simple work of making dictionaries cheaply ($2 per hour) and in large quantities.
https://t.co/GOvhS1gdJg

[Tetsuya @_HaneAI]

#ChatGPT is heavily based on this paper that Google researcher published in 2017.
Due to a masterful paper published by them, google now has to declare a Code Red emergency and respond to machine learning tools such as ChatGPT.

"Attention Is All You Need"
https://t.co/G4j4UzrlnQ

[Tetsuya @_HaneAI]

The technical achievement of #ChatGPT has been to create a dictionary with many key, values and train it to the machine learning model presented in this paper.

[Tetsuya @_HaneAI]

To paraphrase the #ChatGPT process in machine learning terms, it is quite powerful because it has a structure that allows setting the target variable in words.

[Tetsuya @_HaneAI]

Finally. This article discusses a college student who was caught using a text generating AI tool to write a paper. The AI tool, #ChatGPT, was not able to produce content that was up to par for the college's standards, leading to the student's suspension. https://t.co/oQwMmJ16EX

[Tetsuya @_HaneAI]

The onion's article is really teasing Elon Musk lol.

[The Onion @TheOnion]

Elon Musk Receives Experimental Neuralink Implant In Attempt To Delete Memory Of Being Booed https://t.co/iL39zKCkQZ

[Tetsuya @_HaneAI]

There was no announcement beyond what was expected, like a minor update. They are strong in manufacturing and their manufacturing strength allows them to enter all industries. It's like they have no enemies.

Neuralink Show & Tell Event (23 Minutes)" https://t.co/4oKOsFdeKT

[Tetsuya @_HaneAI]

I feel like I can work 120 hours a week. I work 17 hours a day eating, sleep 6 hours, and have 1 hour left.

SCARY WORK ETHIC - Elon Musk Motivational Video https://t.co/D9zXPrTtpN via @YouTube