Yanming Di

10.3K posts

Yanming Di

Yanming Di

@diydatascience

Be positive. Retweets/likes ≠ endorsements, just bookmarks. Opinions are OP's alone. My political posts are satirical—I’m not responsible if you’re offended.

Katılım Ağustos 2017
1.3K Takip Edilen373 Takipçiler
Sabitlenmiş Tweet
Yanming Di
Yanming Di@diydatascience·
@Windows Where should I send a complaint that the Windows 11 UI is so much worse than Windows 10? It's basically a disaster. Why would a company spend so much money to make things worse?!
English
3
0
17
4.3K
Yanming Di
Yanming Di@diydatascience·
@hbkazemi33 @ValerioCapraro I don't think we'll give decision-making authority to AI either. People may base their decisions on AI recommendations, but that's not fundamentally different from making "data-driven" or "evidence-based" decisions?
English
0
0
0
8
H.Kazemi
H.Kazemi@hbkazemi33·
@diydatascience @ValerioCapraro Big difference between tools such as computers and AI. We never gave computers the ability to make critical decisions based on vague rules. Yes, AI will be orders of magnitude more impactful - good and potentially bad.
English
1
0
0
15
Valerio Capraro
Valerio Capraro@ValerioCapraro·
The Pope is making exactly our point. LLMs “may imitate or even simulate, but they do not understand.” This is the core epistemic fault line. Most AI evaluation is still based on one assumption: if a system statistically approximates human behaviour, then it is close to human intelligence. But approximation is not intelligence. Simulation is not understanding. LLMs can produce the right answer without knowing why it is right. They can simulate empathy without feeling. They can imitate judgment without responsibility. They can generate coherent explanations without having a world to which those explanations are accountable. Stop confusing behavioural similarity with cognitive equivalence. Human understanding is embodied, affective, relational, motivational, and normative. It is not just the production of plausible text. * Full paper in the first reply
Valerio Capraro tweet media
English
167
734
2.8K
91.2K
Yanming Di
Yanming Di@diydatascience·
@VesperAegis @ValerioCapraro That is understandable and more pratical, but just like other engineering problems: you clearly define the goals and see whether you can meet them ... The term AGI itself is not much meaningful, no one really knows what it means
Corvallis, OR 🇺🇸 English
1
0
0
17
Vesper Aegis🇵🇸
Vesper Aegis🇵🇸@VesperAegis·
There's no reason to really take this as antagonistic on either side. If one's goal is to create AGI, you should practically take those points into consideration: Embodiment, experience, and moving beyond a predictive/tokenized model towards something that holistically understands a total set of knowledge and interconnected causal inference is necessary for the creation of consciousness and sentience.
English
0
0
2
59
Yanming Di
Yanming Di@diydatascience·
@business Do you have data supporting your claims or simply speculating ... ?
English
0
0
0
41
Yanming Di
Yanming Di@diydatascience·
@ewarren btw AI was not built by you, it is by the people you wanted to destroy
Corvallis, OR 🇺🇸 English
0
0
0
4
Elizabeth Warren
Elizabeth Warren@ewarren·
AI was built on your creativity, your tax dollars, and your electric grid. The profits shouldn't only flow to billionaires. It's time to tax AI and invest in people.
Elizabeth Warren tweet media
English
560
415
1.4K
100.7K
Yanming Di
Yanming Di@diydatascience·
@ronrule Yeah, with inflation high we will all need to millionares just to get by ...
Corvallis, OR 🇺🇸 English
0
0
0
12
Ron Rule
Ron Rule@ronrule·
I’m going to say something controversial… A college education is no longer worth the time or cost. AI is going to make more millionaires in the next five years than we’ve ever had in human history. The only question is whether or not you’re going to be one of them. If you’re about to finish high school, your time would honestly be better spent taking free Claude courses on Skilljar than spending one second inside the walls of a University. It’s one thing if you plan on becoming a Doctor or going into another regulated/gatekept field, but outside of that, the only thing you’re going to get out of college is a lot of debt and an employee discount at Starbucks. I’m saying this because I see a pattern forming. It reminds me of the tech boom of the late 90’s, but for everything. Back then, tech certifications like A+, MCSE, Cisco CCNA, and other vendor certifications held WAY more value than a Computer Science degree. The tech was changing so rapidly that by the time the guys who went for a CS degree got out, they were four years behind. Meanwhile those of us who went the certification route, or just KNEW the tech, had already scooped up all the good jobs and formed defined career paths. This is going to happen again, but for ALL sectors. The guy who can build AI automation flows will have endless opportunities. Employers will be hiring fewer and fewer people with degrees, and more AI experts to REPLACE the people with degrees they already have. It won’t be because of greed or profits. It will be for survival. AI is an equalizer. It levels the playing field between an established business with thousands of employees and a startup competitor with 1-10. Companies with massive overhead and personnel are going to be competing against new companies with NONE of their costs. This has happened before. Many of you witnessed it first hand. Retailers that had been solid for 50+ years and didn’t embrace ecommerce fast enough didn’t survive. They lost too much market share to online sellers who had none of their operational overhead. AI brings that to every industry. Their choice will be to embrace AI and survive, or die. In order to survive, they’re going to be hiring AI and automation experts. Lots of them. You need to be one of those guys. Because nobody is going to give a shit about your degree. And the best part is if you become one of those guys, and STILL can’t find a job, you’ll have the expertise needed to become the competitor they fear. There is literally no risk or downside to learning this. It’s all upside. There is MASSIVE risk to betting on college.
English
20
1
17
1.2K
Yanming Di
Yanming Di@diydatascience·
@ewarren besides tax and lie, what else can you do?
Corvallis, OR 🇺🇸 English
0
0
0
12
Dudes Posting Their W’s
Dudes Posting Their W’s@DudespostingWs·
This is what the older generations should try and do more often
Dudes Posting Their W’s tweet media
English
1.6K
16.2K
207.6K
3.3M
Yanming Di
Yanming Di@diydatascience·
@Kekius_Sage The debate is moot: what is important now is how you best use AI
Corvallis, OR 🇺🇸 English
0
0
1
19
Kekius Maximus
Kekius Maximus@Kekius_Sage·
Elon Musk just settled the biggest AGI debate in tech. Yann LeCun said: "Even humans don't have general intelligence. We evolved to specialize, not to be universal." Demis Hassabis said: "You're confusing AGI with universal intelligence. The human brain is a biological version of the Turing machine, and AI can absolutely reach that level of generality." Elon Musk affirmed: "Demis is right." One claims AI will be as smart as humans, while the other argues humans themselves aren't necessarily that smart. Who is right?
Kekius Maximus tweet mediaKekius Maximus tweet media
English
315
60
635
123K
Yanming Di
Yanming Di@diydatascience·
@Kekius_Sage I am sure we are not Turing Machines
Corvallis, OR 🇺🇸 English
0
0
0
11
Yanming Di
Yanming Di@diydatascience·
@YaleEconomics @nberpubs That's why you have national standardized tests: your peers are no longer your main competitors. There could be a distinction between qualifying exams (such as Common Core assessments) and competitive exams (such as the IMO).
English
0
0
0
15
Yale Department of Economics
"This paper examines the theoretical and empirical consequences of rank-based reward systems in schools in which students’ performance and effort are evaluated relative to their peers." NEW in @nberpubs by Mark R. Rosenzweig & Bing Xu: nber.org/papers/w31135
Yale Department of Economics tweet media
English
1
2
19
1.8K
Yanming Di retweetledi
unusual_whales
unusual_whales@unusual_whales·
Workers should see AI as a powerful tool that amplifies their abilities rather than as a threat to their jobs, Jeff Bezos has said, per YF
English
210
47
662
136.3K
Adam Ozimek
Adam Ozimek@ModeledBehavior·
It's true that unemployment has gone up for recent college graduates. Many are concerned that AI is doing this. But how then do you explain the simultaneous rise in teenager unemployment? They work in restaurants and ice cream shops. Does that sound like AI? agglomerations.eig.org/p/smells-like-…
Adam Ozimek tweet media
English
18
21
93
8K
Bloomberg
Bloomberg@business·
JD.com founder Liu Qiangdong vowed to prevent the e-commerce firm’s 900,000-strong workforce from losing their jobs to automation as fears grow over the adoption of AI and robotics bloomberg.com/news/articles/…
English
8
7
47
15.4K
Yanming Di
Yanming Di@diydatascience·
@kearney_melissa On a related/unrelated note: for the same reason, p-value will make a comeback too 😂🤣😅
English
0
0
0
84
Melissa S. Kearney
Melissa S. Kearney@kearney_melissa·
A commitment to academic excellence and objective metrics is making a comeback! 👏 Neither obscuring academic preparedness nor eliminating metrics that reveal socioeconomic gaps was ever going to be the right way to expand access to highly selective universities. My bet: The backlash against test optional and grade inflation is going to make things harder for admissions officers and varsity coaches, and easier for teaching faculty.
Steve McGuire@sfmcguire79

Yale fully reinstates its pre-pandemic SAT/ACT requirement in admissions. “These test scores are strong predictors of a student’s future Yale academic performance, and there is evidence that they are less subject to bias than other elements of an application.”

English
7
10
93
13K