OpenAI releases o1, its fir=
st model with =E2=80=98reasoning' abilities (8 minute read)
OpenAI has released o1 and o1-mini, the=
first in a series of reasoning models that have been trained to answer mor=
e complex questions faster than a human can. The model is better at writing=
code and solving multistep problems than previous models, but it is more e=
xpensive for developers and slower to use than GPT-4o. The release is still=
in preview to indicate how nascent it is. ChatGPT Plus and Team users shou=
ld already have access to the model, while Enterprise and Edu users will ge=
t access early next week. OpenAI plans to bring o1-mini access to all free =
users, but it hasn't set a release date yet.
|
SpaceX Astronauts Conduct S=
pacewalk, Putting New Spacesuits to Test (5 minute read)
The private-citizen astronauts travelin=
g with SpaceX have completed the first-ever commercial spacewalk. The space=
walk kicked off shortly after 6 am ET on Thursday. The livestream showed te=
ch industry billionaire Jared Isaacman leaving the SpaceX Crew Dragon and c=
onducting a series of mobility tests before coming back after around an hou=
r. The entire spacecraft was exposed to the vacuum of space during the spac=
ewalk. The crew members were wearing SpaceX-designed and manufactured suits=
that were previously untested in orbit. A clip from the livestream is avai=
lable in the article.
|
|
Science & Futuristic Technology
|
Face to face with Figure's =
new humanoid robot (6 minute read)
Figure's humanoid robots were put to wo=
rk at BMW's Spartanburg, South Carolina plant for a couple of weeks earlier=
this year - they will be returning for good in January. The automotive ind=
ustry has been eager to adopt the humanoid robot form factor, which is in l=
ine with its deployment of robots on factory floors. Figure aims to hit a s=
ub-$20,000 price point, but it will take some time. In the meanwhile, the c=
ompany is testing its robot's efficacy in helping out with tasks around the=
home. This article features a video of Figure's second-generation robot up=
close.
|
Google DeepMind teaches a r=
obot to autonomously tie its shoes and fix fellow robots (1 minute read)
Google DeepMind researchers recently re=
leased a paper showcasing a method for teaching robots to perform a range o=
f dexterous tasks. It used a new learning platform called ALOHA Unleashed p=
aired with a simulation program, DemoStart, to teach robots by watching hum=
ans. The research showcased how systems can learn to perform complex, dexte=
rous tasks from a visual demonstration. Videos of robots trained with the m=
ethod performing tasks are available in the article.
|
|
Programming, Design & Data Science
|
Your Next Store (GitHub Rep=
o)
Your Next Store is a modern ecommerce s=
ite template in Node.js that uses Stripe as the backend. It gets all of the=
products, prices, descriptions, and categories from Stripe, making it easy=
for users who know Stripe already to use. Your Next Store supports simple =
product variants, Stripe Webhooks, Stripe Tax, and more. A demo site and vi=
deo are available.
|
How GitHub taught me to Mic=
romanage (8 minute read)
Good feedback cultivates quality work a=
nd professional growth - bad feedback degrades quality and erodes relations=
hips. Bad feedback makes critiques but doesn't offer a better direction. OK=
feedback provides direction. Good feedback educates to produce a better re=
sult. All of this depends on context - detailed feedback may be welcomed in=
circles where people are trying to improve others around them, but it can =
be toxic in environments where few people care about craft.
|
|
The Success of Black Myth: =
Wukong And What It Means for Media (8 minute read)
Black Myth Wukong is one of the fastest=
-selling games of all time. The game cost $70 million to make and has gross=
ed close to $1 billion so far. The rise of regionally rooted AAA video game=
s has been slower than in film due to their comparatively high production c=
ost, the rate of technological, creative, and change, and a smaller market.=
The growing capability and ease of use of game engines as well as the affo=
rdability and size of asset store libraries will result in a growing number=
of hits that target non-Western regions, countries, and cultures, some of =
which will go far beyond those markets.
|
The Button Problem of AI (1=
2 minute read)
Productivity software tends to buttonif=
y AI into small, discrete tasks because most models are currently incapable=
of building and following complex plans as independent agents. Project bui=
lders are forced to bind models to tasks that are small and easily definabl=
e because they aren't able to handle knowledge-labor workflows in their tot=
ality. Simply replacing small portions of a workflow with a button isn't en=
ough to meaningfully change the work itself. The buttonification of AI is a=
reminder that we are still in the 'training wheels' phase of the supposed =
AI revolution.
|
|
|
Congrats to Ben Stef
for referring 5 friends to TLDR. Thanks for helping us spread the word=
! =F0=9F=99=8F
Love TLDR? Tell your friends and get rewards!
|
|
Share your referral link below with friends to get free TLDR swag!
|
|
|
|
| Trac=
k your referrals here. |
|
Want to advertise in TLDR? =F0=9F=93=B0
|
If your company is interested in reaching an audience of tech executives, d=
ecision-makers and engineers, you may want to advertise with us.
If you have any comments or feedback, just respond to this email!
Thanks for reading,
Dan Ni & Stephen Flanders
|
|
|
|
![3D"Vanta"](3D"https://images.tldr.tech/vant=)