Dorothea Baur (Dr.)
@DorotheaBaur
Followers
9,395
Following
5,095
Media
2,370
Statuses
23,140
Ethics in tech & finance: AI ethics, ESG, CSR, sustainability. Expert, consultant, speaker, lecturer, author. To the point. @dorotheabaur .bsky.social
Zürich, Schweiz
Joined February 2018
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Padres
• 105401 Tweets
Satoshi
• 96706 Tweets
Dodgers
• 84790 Tweets
#究極のポテチ計画
• 29003 Tweets
ドジャース
• 28395 Tweets
Ghar Vapasi
• 26444 Tweets
シロネン
• 25864 Tweets
KAIRAIN FIRST BILLBOARD
• 21890 Tweets
Millonarios
• 18546 Tweets
Teoscar
• 16777 Tweets
プレミア12
• 14218 Tweets
Rojas
• 13894 Tweets
Buehler
• 12915 Tweets
侍ジャパン
• 12256 Tweets
Mookie
• 10857 Tweets
Peter Todd
• 10464 Tweets
Pinned Tweet
Wer heute fliegt, befindet sich guter Gesellschaft. Divers, politisch und sozial heterogen; und kreativ, wenn es um Ausreden geht. Die Chance ist gross, dass auch Sie in den Argumenten Ihrer Mitreisenden moralische Entlastung finden, schreibe ich
@NZZ
.
10
9
60
Both Roger Federer and I can play tennis.
Both the vaccinated and unvaccinated can spread the virus.
4K
1K
7K
4
133
947
A single bitcoin transaction uses as much power as "the average US household consumes in a month, and is responsible for roughly a million times more
#carbon
emissions than a single Visa transaction". What can I say?🤷♀️
#climatechange
4
395
652
"The main use case for LLMs is writing text nobody wanted to read. The other use case is summarizing text nobody wanted to read. Except they don’t do that either"
Touché 😅
@davidgerard
ht
@ahcastor
1/2
7
127
460
"For the first time in history developing brains are being fed choppy low-grade and cheaply-produced synthetic data created en masse by generative
#AI
, instead of being fed with real human culture. No one knows the effects, and no one appears to care" ➡️
18
174
436
Mit
@schneimere
hat X / Twitter eine der witzigsten, authentischsten und differenziertesten Schweizer Politikerinnen gesperrt. Und all das wegen einem ironischen Tweet, den Andreas Glarner vollkommen verdreht hat. Shame on you, X!
#bringbackMeret
105
61
380
Believing that human-level intelligence will be "attained simply by making systems excel at the imitation game", "would be like imagining that if we can make an aeroplane fly faster and faster, eventually it will lay an egg".
#AI
#RebootingAI
ht
@dmonett
15
101
285
„Wer aus politischen Gründen lieber Vergewaltiger verteidigt, als das Leiden missbrauchter Frauen anzuerkennen“, braucht einen Realitätscheck. „Wer die Entrechtung realer Frauen aufgrund imaginierter Opferhierarchien“ nicht erkennt, „ist falsch abgebogen“.
13
47
244
Brilliant text from
@Tyler_A_Harper
on the danger of AI making us less human. AI "could significantly diminish humanity, even if machines never ascend to superintelligence, by sapping the ability of human beings to do human things". 1/6
14
57
240
Today a law professor got called out for including fake references and quotes in a (non-peer reviewed) article.
The obvious question was: Did she use gen AI for it?
The incident is layered with irony for several reasons:
Thread 1/4 ➡️
8
70
219
Behind seemingly automated systems are "large amounts of low paid labour".
#AI
"is neither artificial nor intelligent. It is made from natural resources". People perform "the tasks to make the systems appear autonomous".
@katecrawford
#sustainability
5
88
202
To all those who struggle to understand the difference between a human who goes to a museum and paints a picture back home and the way generative AI works - maybe this helps?
Unless your lack of understanding is ideologically motivated of course...
#genAI
#AI
15
52
206
Long but great piece
@techreview
. If you have time, read it (takes ca. 1h). Here are some of my favorite quotes. "AI has come to mean all things to all people... It can feel as if different camps are talking past one another, not always in good faith" 1/12
1
41
196
Only in Switzerland 🇨🇭: Ein Wegweiser, der zu weiteren Wegweisern verweist. In 10 Minuten!
26
10
192
"There will never be enough computing power to create
#AGI
using machine learning that can do the same [as humans], because we’d run out of natural resources long before we’d even get close".
Must definitely read full paper from
@o_guest
@IrisVanRooij
!
15
95
207
"Mama, war ich eigentlich ein Wunschkind?"
"Nein, du warst ein Rentenkind."
15
20
174
@sybileggarter
Nein. Aber ich kann dir erklären, was mich bewegt, wenn ich aus einem Auto auf der Autobahn heraus Menschen sehe, die von der Brücke auf den Verkehr starren: Panik.
6
0
176
Dass eine Partei, die das Label "liberal" im Namen trägt, wegen einem blöden Insta-Post (gelöscht und von Bitte um Entschuldigung gefolgt) ein Parteiausschlussverfahren gegen ein Mitglied einleitet, ist ein Armutszeugnis.
59
13
177
Auch die "liberale" NZZ wird zur schadenfreudigen Befürworterin der Cancel Culture, gegen die sie sonst täglich wettert.
- "Jetzt bekommt Ameti ein paar Schlücke ihrer eigenen Medizin verabreicht"
- Die Parteiausschlussforderung nennt sie "konsequent"
93
15
173
Strong metaphor: Big Dick Data projects are „are characterised by masculine fantasies of world domination. [They] fetishise large size and prioritise it, along with speed, over quality, ignore context and inflate their technical capabilities“ Ht
@vavacolor
9
70
174
Human thought is always embodied & enacted. Data used to train LLM's can't "capture the motivational, participatory, and... social aspects that ground meaning making by people".
@Abebab
on "Mistaking engineering Achievements for Human Linguistic Agency"
3
65
176
"Democracies should never allow businesses to grow so powerful that they can defy governments" & courts. Otherwise they "can distort public discourse, sway elections, evade taxes, and deepen societal divides"
"a direct threat to democracy"
@francesca_bria
12
90
173
It took me almost half a century until I got to give my first guest lecture at ETH Zürich. But today is the day 🤩
15
2
166
"Each algorithm cited in the registry lists datasets used to train a model, a description of how an algorithm is used, how humans utilize the prediction, and how
#algorithms
were assessed for potential
#bias
or risks" great initiative enabled by
@ai_saidot
3
73
162
Law enforcement uses
#AI
that turns
#DNA
into a face and then run the "face" through
#facialrecognition
software, thereby mixing and matching technologies "in unintended ways, using untested algorithms to single out suspects based on unknowable criteria"🤔
13
61
158
Social media algorithms, are typically optimized for engagement.
@StanfordHAI
researchers finds if this function is replaced with "a societal objective function" instead, that optimizes for democratic values, partisan animosity decreases.
4
50
155
"AI may also help managers directly. Its capabilities in empathy, summarisation, and customisation make it a powerful tool for coaching and mentoring"
Empathy? You must be kidding. AI has about as little empathy as any other corporate psychopath.
27
24
154
Zug verzichtet aus Nachhaltigkeitsgründen auf Feuerwerk. Kein Problem: Bitcoin Millionär übernimmt. Geld verbrennen und Umwelt belasten ist schliesslich seine Kernkompetenz. Armes reiches Zug!
41
16
143
The promised benefits of
#AI
will fade just like the promises "of the metaverse, the crypto industry" etc. "But the harmful uses of [AI] will stick around, unless... action is taken to stop [them] from lingering long after the bubble bursts"
@parismarx
9
49
148
I just won the biggest mandate in my career (
#ESG
in the financial industry). The best thing is their reason for choosing me: "We want you because you are 'cheeky' in a positive way, and you don't shy away from controversies". I am over the moon.
#authenticity
rules.
#ethics
😊
15
2
143
Nein, natürlich hat die Kritik an Harris’ Lachen nichts mit ihrem Geschlecht zu tun… Aber wenn sie aufs Lachen verzichtet, wirkt sie „staatsmännischer“ (sic!).
🤦♀️🤦♀️🤦♀️
via
@NZZ
68
8
145
Hotel rooms with curtains that don’t close properly? Glad I remembered that life hack I saw here some years ago.
6
8
139
5
8
137
"AI is made in our image, and reflects the dominant value system of today. What we should be concerned about is how AI blocks our capacity to reinvent our values, 'preventing us from knowing how to make a future at all'."
On
@ShannonVallor
's new book.
6
52
136
#NHS
gives
#Palantir
access to
#data
including "details of political and religious affiliation and past criminal offenses" Can anyone explain to me how past criminal offenses could be relevant in dealing with the
#COVID19
pandemic? 🤔 ht
@SophieTaysom
5
137
129
"Wherever
#algorithms
have been used for justice [e.g. welfare payments, prison sentencing], these turned out to be gender and racially biased. If you can’t demonstrate fully and test that an algorithm isn’t biased, then it should not be used"
@NoelSharkey
6
61
124
Must read
@CT_Bergstrom
's new book: "technocratic worship of
#data
", "tyranny of numbers"; people use "data to bullshit" when presenting claims in an entirely quantitative way. If someone uses
#AI
for
#phrenology
, "you need to be able to say that is nuts".
3
49
130
No, running a series of different prompts through GPT-4 & remixing the results (i.e. synthetic text) isn't a valid empirical method for making predictions about whatever subject you're interested in.
Thnx for clarifying
@emilymbender
@mikarv
&
@alexhanna
4
36
125
Basilikummer. Das Gefühl, das sich ausbreitet, wenn jeder Basilikum innerhalb von 3 Tagen verwelkt.
43
1
125
A crypto-mining company in Pennsylvania is seeking to burn tires to produce bitcoin, risking creating "highly toxic chemicals linked to cancer and known to be slow to break down in the environment".
Yeah, bitcoin is definitely not just hot air. It's hot, dirty, toxic air.
> "The company also says it sometimes needs to burn tire-derived fuel – made of shredded vehicle tires – to make the combustion of waste-coal more efficient."
Can't make this stuff up. It's literally a tire fire to trade imaginary magic beans.
9
102
251
1
32
120
"limiting training data to public domain books... created more than a century ago.. would not provide AI systems that meet the needs of today’s citizens".
Nice try by
#OpenAI
to manipulate us by appealing to "needs", which we didn't even know we had.
9
26
120
Happy about my hard copy of „The ethical
#algorithm
“ by Michael Kearns & Aaron Roth. A great book for anyone interested in how much
#ethics
can be ‚encoded‘ and what role human judgment plays. Very helpful also for my
#AIethics
teaching!
#privacy
#fairness
8
32
121
Ich starte den Tag mit einer hemmungslosen Gefühlsoffenbarung.
Frauen lachen und Männer hoffentlich auch!
Schönen Sonntag!
16
4
120
I had a similar experience recently: I was on a panel, as the only woman. The moderator introduced all the men with their titles and expertise. All he said about me was "Please also welcome Dorothea Baur. She is also interested in AI".
😂😂😂.... 😭😭😭.... 😵💫😵💫😵💫
I was on a panel the other day where the moderator presented the other (male) panelists with all of their titles ("how do you fit all that on a business card?") and just...forgot?... to present me by anything other than my name.
12
15
189
20
20
122
In
#tech
"there’s a quantitative fallacy that if you can’t do maths on it, you can just ignore it. And so you just incinerate the qualitative elements and do maths on the dubious quantitative residue that remains" ht & feat.
@mariafarrell
#machinelearning
2
49
119
Excellent primer on the environmental impact of
#AI
by
@SashaMTL
@mmitchell_ai
et al., which looks beyond the energy consumption of training AI models but instead covers the entire lifecycle of AI and includes other resources like water and minerals.
2
39
121
Die Klimaberichterstattung der
@NZZ
erreicht langsam Weltwoche-Niveau. Ich überlege mir ernsthaft, ob ich unter diesen Umständen mein Abo erneuern soll. Rechtspopulistische Propaganda mit schmissigen Alliterationen ("Klima-Kommunisten", haha!) bringt null Erkenntnis, nur Ärger.
55
9
115
Ist das wirklich eine smarte Idee von der Helvetia Versicherung? Bullshit-generating
#ChatGPT
mit dem eigenen
#Chatbot
verknüpfen und den Raum öffnen für falsche Antworten auf triviale, aber für Kunden zentrale Fragen?
#LLM
#insurance
@tippinpointnews
16
16
116
Trustworthy AI is a misnomer. "Trust is a way of delegating a task and no longer .... supervising the way the task is being performed". It's dangerous to "trust and forget about AI". We should make
#AI
reliable rather than trustworthy. says
@RosariaTaddeo
6
42
114
"The meaning of a word or sentence is intimately related to the human body: people’s abilities to act, to perceive and to have emotions. Human cognition is empowered by being embodied". Without body,
#LLMs
"will never understand language the way humans do"
9
27
110
#LLMs
"produce untrustworthy output, which is then sucked back into other LLMs. The models become permanently contaminated. Scientific journals’ peer-review processes will be overwhelmed, 'leading to a precipitous drop in reputation'."
@GaryMarcus
@FT
5
32
112
I have finally installed
#SignalApp
but I am really not happy with the fact that apparently everyone who is on it already and who has my number, gets notified about that. Even contacts I have long deleted from my phone. Is this
#privacy
-friendly? Does anyone share my anger?
29
8
106
😳Und was ist mit Ethnologie und Blockchain, Gender Studies und Blockchain, Psychologie und Blockchain, Klimawissenschaften und Blockchain… etc?
34
6
109
The current
#AI
discourse fails "to appreciate why it is so important to champion human dignity. There is risk of creating a world in which meaning and value are stripped from human life."
@carnegiecouncil
@WendellWallach
7
35
106
Even if it was somehow possible to establish a scientific correlation betw
#physiognomy
&
#criminality
& even if some ‚miracle‘ made it possible to totally de-bias datasets, this research is unethical. It‘s not an empirical, but a moral question.
#AIethics
4
45
107
I'm reading this and think of the guy who (in a professional context) called me "ignorant and incompetent" earlier this year when I pointed out the environmental costs of
#bitcoin
. I hope he reads this, but I doubt he does.
#cryptocurrencies
#CO2
#ethics
10
39
102
My cheeks hurt from smiling😊 I've won an important consulting job. I was so scared of pitching against 'big players'. But client says they feel like I really listen to their concerns and that I'm very professional and passionate. I'm the happiest one-woman-entrepeneur on earth!
13
0
105
Ok, here's my first TV appearance ever... in a debate on
@CNNMoneyCH
together with Heinz Rothacher (Complementa) on the state of
#ESG
investing among Swiss pension funds...I really enjoyed the debate!
#sustainablefinance
#sustainability
#climatechange
#SRI
11
11
101
#AI
skeptics "are begging Silicon Valley to take a beat before unleashing AI to the world. But tech companies, faced with the most powerful computing innovation in a generation, are running around like kids who just found their dad’s gun".
6
39
104
10
1
103
Synthetic media means: "the people doing some of the jobs we’d most want done by fellow humans ...- the creative work of writing and art that enriches human society - are the first to be targeted by people who seem most divorced from the human condition"
2
31
103
On AI music: "the coming war is not really one between human and machine creativity; the two will forever be incommensurable. Rather, it is a struggle over how art and human labor are valued—and who has the power to make that appraisal"
@matteo_wong
1/2
AI Can’t Make Music
But that doesn’t mean it poses an empty threat to musicians.
www.theatlantic.com
5
42
98
I'm very happy to join Women in AI
#Switzerland
as their ethics lead and support their mission to increase
#diversity
and
#inclusion
in
#AI
. We've only just allowed women to vote here, now we already want them to do
#AI
! How crazy is that? 😜
#AIethics
12
16
96
"When it comes to disinformation, 'social media took the cost of distribution to zero, and generative AI takes the cost of generation to zero'"
@noupside
.
Even worse, personalised disinformation can be "generated to appeal to small groups".
ht
@mario_gug
4
42
97
A masterpiece of
#data
journalism: interactive illustration of an allegedly 'risk scoring' but “essentially random guessing” (
@mmitchell_ai
) and discriminating algorithm used to identify welfare recipients suspected of committing fraud in Rotterdam.
#AI
4
44
94
If an
#algorithm
performs well when tested on CoNLL-2003, then it’s viewed as successful. But in reality, this 'success' "means the model has been built to reinforce some of the dataset’s initial
#bias
" Great illustrative piece on
#AIethics
by
@haydenfield
3
43
95
I made it! I'm
a) called an "intellectual",
b) by a crypto
#Bitcoin
Maximalist,
c) in a piece about the great
@concodanomics
c) together with
@Bitfinexed
@CasPiancey
@amycastor
@smdiehl
d) my name is misspelled as that of an ex German minister
🤡
13
15
96
Ernste Frage: warum spricht man eigentlich immer von der Anzahl Kinder pro Frau und nie pro Mann? Warum sagt man "Frauen bringen nur noch x Kinder zur Welt" und nicht "Männer zeugen nur noch x Kinder"? Weil man bei Männern nicht die gleiche Kontrolle hat?
65
6
94
Liebe ältere
#Follower
: wer von euch erinnert sich noch an den ABM (Warenhaus) ? Wir scherzten ja immer, dass die Buchstaben für "Alles billiger Mist" stehen. Aber dieses Handtuch begleitet mich seit 1988. Nun nur noch zum Sport, dort aber tadellos. Bin entzückt!
#Nachhaltigkeit
13
1
93
Absolute must read by
@anilkseth
who argues that, we should not try to build conscious
#AI
(no matter how far away we are from it), and why our tendency to (wrongly) attribute conciousness to
#AI
contorts our ethical attitudes. 1/4 ht
@GaryMarcus
10
24
92
@NC_Renic
No worries. The only insecurity you should have in your interaction with Deutsche Bahn is whether it will ever get you to your destination. And I don't think they differentiate between academic titles in their miserable service.
2
1
91
Great piece! Without substantive
#AI
regulation "the foibles and idiosyncrasies of its creators take on outsized importance".
"The
#AI
that gets built relies on a handful of people who built it", and these people have disproportionate impact.
@rayidghani
12
34
91
Great thread on why
#cryptocurrencies
like bitcoin are Ponzi schemes - notably in contrast to fiat money, insurances, the stock market etc.
Since crypto has hit the mainstream, we've witnessed an overload of "Ponzi whataboutisms", attempts to justify crypto Ponzi schemes by comparing them to various other "things".
There's an easy way to identify a Ponzi scheme...
74
832
3K
0
49
83
I feel very honored to be listed as one of 100
#BrilliantWomeninAI
! Amazing initiative by
@MiaD
! I take it not only as a recognition of past work but even more so as an invitation or rather *obligation* to remain persistent... sorry guys😄
#AIEthics
8
16
91
Again, dystopian use of
#AI
in India:
#algorithms
declare people dead and stop paying them
#pensions
. Crazy story of what happens when people are ruled by an interlocking web of flawed databases with the absurd claim to be the "single source of truth".
3
57
90
Must read by
@yudapearl
: Researchers convinced "that all knowledge comes from the
#data
" struggle to accept an "alien paradigm, according to which extra-data information is needed, in the form of man-made, causal models of reality"
#AI
ht
@FrankPasquale
6
25
89
Great piece on
#crypto
's false assumptions about creation of money, about bookkeeping and why with "the crypto industry largely living outside the fundamental laws of finance and economics", auditors have a hard time doing their job.
@MartinCWWalker
1/4
3
25
88
Some people use
#AI
to create deep fakes, plagiarize the work of others, and contaminate the internet with stochastic bullshit. Others use it to develop a "potentially powerful new approach to cancer detection, monitoring and treatment".
Ht
@mireillemoret
3
16
89
Must read! "talking about ethics, morality, and values stimulates people in a way that’s different from other kinds of business work". Like a"fight-or-flight response" Couldn't agree more,
@mmitchell_ai
!
Also featuring
@baxterkb
&
@timnitGebru
#AIethics
1
41
90
Whew, number 19 among influencers in
#AIethics
worldwide! Let me thank the
#algorithm
behind the ranking, which is hopefully free of
#bias
, respects
#privacy
, has a human in the loop - in short, conforms to the highest standards of
#AIethics
😉
9
12
90
True words from
@mer__edith
on whether
#AI
will wipe out humanity: "the data centers can flood as the climate encroaches, we can run out of the water to cool the data centers" etc. But I hope we don't have to rely on
#climatechange
to prevent
#AGI
🙈 1/2
3
39
89
There's so much talk & confusion surrounding
#audits
on
#AI
,
#algorithms
etc. Here's a very (!) helpful taxonomy distinguishing independent audits from internal audits assurance, consulting, assessments. by
@ForHumanity_Org
&
@sheabrownethics
#AIethics
4
39
85
I don't agree with the notion of algorithms having agency. When we attribute agency to algorithms, we lay the foundation for viewing them as actors with responsibilities (rights and duties). Just because they have an impact, doesn't mean they have agency 🤷♀️
10
18
86
Ich wurde gerade erstmals von einem Journalisten kontaktiert, der mich auf gefunden hat. Danke,
@alliance_F
, für diese tolle Plattform! Liebe Mit-Expertinnen, schaltet euer Profil dort auf. Wir wollen euch hören und lesen!
#sheknows
0
8
84
'Invitation': "Dear Dr. Baur, we'd like to have you as a keynote speaker at our event. it will cost you xxxx CHF"
My answer: "thanks, but I don’t believe in the concept of paying for my own work. I do however strongly believe in getting paid for my work by others"
#notforfree
6
5
86
Anything meaningful in writing comes from the "nerve-shredding, fingernail-yanking uncertainty of writing". To have AI strip all that away, would be to render us "wordless, thoughtless, self-less. Give me the shredded nerves and yanked fingernails any day"
10
25
85
School canteens introduce
#facialrecognition
for a totally mundane, riskfree task, arguing that it is "more
#Covid
-secure", and promising a 'frictionless experience'. What's next?
#voicerecognition
for ordering Pizza?
#DNA
tests for opening mailboxes?
9
42
84
Couldn't disagree more with this piece's view on
#AI
in
#recruiting
, on the 'non-value' of qualitative interviews, on the pseudo-measurability of every character trait, on
#AI
being the solution to overcome all
#bias
etc
#emotionrecognition
HT
@RobMcCargow
9
28
81
This may be an impressive step in
#robotics
but I struggle to acknowledge what's described here even as a "primitive form of empathy“. Empathy goes far beyond predicting movements. It's about understanding/sharing the feelings of another. ht
@MedAIActivist
9
31
84
Happy new year to everyone!
My only resolution is to hang in there.
Peace! 🕊️
10
0
84
Bin ich die einzige, die bewusst keine Videos von Bidens Aussetzern und Stolperern schaut - und zwar aus einer Mischung von Mitleid, Fremdscham und Verzweiflung?
Ich kann und will das einfach nicht sehen.
29
0
82
Good table on different tools assessing
#Algorithms
, making key distinctions between
#audits
and impact
#assessments
. Import contribution to mitigate confusion. Report by
@AdaLovelaceInst
(from Apr 2020) h/t
@AIEthicsWatch
#AIethics
There's so much talk & confusion surrounding
#audits
on
#AI
,
#algorithms
etc. Here's a very (!) helpful taxonomy distinguishing independent audits from internal audits assurance, consulting, assessments. by
@ForHumanity_Org
&
@sheabrownethics
#AIethics
4
39
85
2
23
81