The Fine-Tuning of the Universe for Life: The Evidence, the Concessions, and What They Mean

Abstract

The universe is fine-tuned for life in the sense that matters for this debate. Across the physics literature, a small set of constants and initial conditions has to fall within narrow ranges, compared with the ranges physical theory allows, for chemistry, stars, planets and observers to exist. Denying this puts a skeptic at odds with most of the physicists who study it, including avowed naturalists. The strongest cases are the following:

The cosmological constant, whose life-permitting value requires a cancellation “accurate to about 120 decimal places” (Weinberg). aish nideffer
The low entropy of the Big Bang, a target Penrose estimated at 1 part in 10^(10^123). ws5
The quark masses and the Higgs vacuum expectation value (vev), which must sit in a small region of a parameter space spanning roughly 60 orders of magnitude (Barr & Khan; Agrawal et al.). osti arxiv
The amplitude of primordial density fluctuations, Q ≈ 2×10⁻⁵. arxiv

Other classic examples are softer. These include the carbon resonance, diproton stability and the expansion rate (which inflation may explain). The best critic, Fred Adams, shows that some parameters can vary more widely than popular accounts claim. stanford Still, the scientists who reject a theistic interpretation mostly explain fine-tuning away rather than explain that it does not exist. Hoyle, Davies, Rees, Hawking, Weinberg, Susskind, Dyson, Penrose and Sean Carroll all accept that there is something to explain; they disagree about what explains it. The honest conclusion is that fine-tuning is an established and robust datum. It is strong evidence that our universe’s life-permitting character is not a coincidence within a single universe. It does not by itself prove God, because the real live competitor is a multiverse with anthropic selection, and that option carries serious philosophical and empirical burdens of its own.

TL;DR

Fine-tuning is real, and serious naturalists concede it. Sean Carroll, the most prominent physicist critic, called it “the best argument that the theists have when it comes to cosmology.” Weinberg called the cosmological constant “remarkably well adjusted in our favor.” Susskind wrote that zeroing “the first 119 decimal places of the vacuum energy” is “most certainly no accident.” The live debate is about the explanation, not the data.
Not every example is equally strong. The cosmological constant, initial entropy, quark masses and Higgs vev, and Q are robust. The carbon resonance, diproton and flatness examples are weaker or contested (Adams 2019; MacDonald & Mullan 2009; inflation). An effective defender leads with the robust cases and concedes the weak ones.
There are four candidate explanations: chance, necessity, multiverse, and design. Chance in a single universe is widely abandoned. Necessity is a promissory note. The multiverse faces the measure problem, Boltzmann brains, the inverse gambler’s fallacy charge and testability objections (Ellis & Silk). Design is a coherent rival. Fine-tuning is evidence that shifts the odds, not a demonstration.

1. What “fine-tuning” claims, and what it does not

Luke Barnes, in the most-cited modern review (“The Fine-Tuning of the Universe for Intelligent Life,” Publications of the Astronomical Society of Australia 29 (2012): 529–564; arXiv:1112.4647), states the thesis carefully: “In the set of possible physics, the subset that permit the evolution of life is very small.” arxiv cambridge He is explicit about what this does not claim. It does not claim that this universe is optimal for life, that carbon-based life is the only possible kind, or that only minor variations on our universe could support life. “These claims, true or false, are simply beside the point.” arxiv

This matters for answering skeptics, because many rebuttals attack claims the literature never makes. Fine-tuning is a counterfactual claim about how physics could have been. It is not an inference from the existence of life to God. In his review Barnes says, “I do not attempt to defend any conclusion based on the fine-tuning of the universe for intelligent life.” arxiv cambridge

The modern scientific literature begins with Brandon Carter, who coined the “anthropic principle” and presented it at the 1973 Kraków Copernicus symposium (published 1974). arxiv It continues through Bernard Carr & Martin Rees (“The anthropic principle and the structure of the physical world,” Nature 278, 1979) and the encyclopedic John Barrow & Frank Tipler, The Anthropic Cosmological Principle (Oxford, 1986). arxiv arxiv Popular syntheses include Martin Rees, Just Six Numbers (1999); Paul Davies, The Goldilocks Enigma (2006; US title Cosmic Jackpot); and Geraint Lewis & Luke Barnes, A Fortunate Universe: Life in a Finely Tuned Cosmos (Cambridge, 2016). The two most important technical reviews are Barnes (2012) and Fred C. Adams, “The Degree of Fine-Tuning in our Universe — and Others,” Physics Reports 807 (2019): 1–111. scispace arxiv Adams is the best-credentialed moderating voice.

Tegmark, Aguirre, Rees and Wilczek (“Dimensionless constants, cosmology, and other dark matters,” Phys. Rev. D 73, 023505, 2006) “identify 31 dimensionless physical constants required by particle physics and cosmology.” They examine 8 anthropic constraints on a 7-dimensional parameter space (α, β, mₚ, ρΛ, Q, ξ, ξ_baryon). The life-permitting region is “boxed-in on all sides” (Barnes’s summary). elsevierpure arxiv

2. The Best Evidence, Ranked by Robustness

Tier 1: The strongest cases

(a) The cosmological constant (vacuum energy). Quantum field theory predicts vacuum-energy contributions vastly larger than the observed value of dark energy, conventionally 10^120 times or more. arxiv +2 Steven Weinberg, a lifelong atheist, wrote in Scientific American (October 1994, “Life in the Universe”): “the existence of life of any kind seems to require a cancellation between different contributions to the vacuum energy, accurate to about 120 decimal places.” aish +2 His technical review (“The Cosmological Constant Problems,” 2000, astro-ph/0005265) states that “the cancellation then has to be accurate to 120 decimal places.” caltech

Weinberg’s 1987 paper (“Anthropic Bound on the Cosmological Constant,” Phys. Rev. Lett. 59: 2607–2610) argued that in a non-recollapsing universe, Λ must not be “so large as to prevent the formation of gravitationally bound states.” semanticscholar +2 This anthropic bound was later seen as a successful anthropic prediction, roughly an order of magnitude above the value measured in 1998. arxiv +2 On the negative side, Barnes notes that the universe recollapses before life can form for ρΛ ≲ −10⁻¹²¹ in Planck units. arxiv Leonard Susskind, The Cosmic Landscape (2005), is blunter: “To make the first 119 decimal places of the vacuum energy zero is most certainly no accident.” Substack +2

Caveat: The “10^120” figure depends on placing the cutoff at the Planck scale. With other cutoffs the mismatch is still dozens of orders of magnitude. Wikipedia The problem shrinks under no reasonable assumption, which is why it is the anchor case.

(b) The low-entropy initial state. In The Emperor’s New Mind (1989, ch. 7, pp. ~339–345), Roger Penrose estimated the phase-space volume available for a universe with our matter content at about 10^(10^123), compared with the volume for an initial state compatible with our observed second law. He concluded that “the Creator’s aim must have been” accurate “to one part in 10^10^123.” ws5 Penrose uses “the Creator” as a figure of speech; he is not a theist and pursues a physical explanation (conformal cyclic cosmology).

Barnes explains why this is not easily dissolved: “Homogeneous, isotropic expansion cannot solve the entropy problem — it is the entropy problem.” arxiv Sean Carroll concedes the datum. In his 2014 debate with William Lane Craig he said that the maximum entropy of the observable universe is ~10^122 while the early universe’s entropy was “enormously smaller,” and added: “I’m not saying there’s not fine-tunings; I’m saying they’re not there for life.” reasonablefaith That last point is a real counterargument (the entropy is lower than life requires), and it is addressed in §4 below.

(c) Quark masses and the Higgs vev (the hierarchy problem). Agrawal, Barr, Donoghue & Seckel (“Anthropic considerations in multiple-domain theories and the scale of electroweak symmetry breaking,” Phys. Rev. Lett. 80: 1822, 1998) scanned the Higgs mass parameter over its full natural range, from +M_P² to −M_P². They found “only for values in a narrow window is life likely to be possible.” osti Values of the vev from the Planck scale down to “about 5 (or perhaps even down to 1.4) times the value in our domain can be excluded,” because nuclei and atoms fail. arxiv Later work narrows the window further: with Yukawa couplings held fixed, Damour & Donoghue (Phys. Rev. D 78, 014014, 2008) found that the Higgs vev “must fall in the range 0.39 < V/V0 < 1.64.” Barr & Khan describe the Higgs mass parameter as “the most highly tuned of the Standard Model parameters, being 10^-34 of its ‘natural’ value.”

Barr & Khan (Phys. Rev. D 76, 045002, 2007) mapped the up- and down-quark masses over a natural range spanning ~60 orders of magnitude. They identified 9 independent catastrophe boundaries, including these:

only a Δ⁺⁺ “helium-like” element;
an unstable deuteron;
hydrogen as the only stable element;
proton capture of electrons destroying atoms;
protons decaying inside nuclei. arxiv

Only a tiny “potentially viable” island survives (reproduced in Barnes 2012, Fig. 2). arxiv This tier also includes the famous fact that the neutron is only slightly (≈1.3 MeV) heavier than the proton. Davies puts it this way: “If it were the other way around, atoms couldn’t exist, because all the protons in the universe would have decayed into neutrons shortly after the big bang” (The Guardian, 25 June 2007). goodreads

(d) The amplitude of primordial fluctuations, Q ≈ 2×10⁻⁵. Martin Rees (Just Six Numbers, 1999): “If Q were smaller than 10⁻⁶, gas would never condense into gravitationally bound structures at all.” If Q were much larger, regions far bigger than galaxies would collapse “into vast black holes,” and “stars would be packed too close together and buffeted too frequently to retain stable planetary systems.” Inflation does not automatically deliver this value. Barnes cites Turok: the inflaton coupling must be adjusted “to be very small, ~10⁻¹³ in Planck units. This is the famous fine-tuning problem of inflation.” Susskind (2005, p. 184, quoted by Barnes) adds: “The various parameters governing the inflating universe must be chosen with great care in order to get the desired result.” arxiv

Tier 2: Real but more contested

(e) Gravity vs. electromagnetism. Rees’s first number, N ≈ 10^36, is the ratio of the electrical force to the gravitational force between two protons; Barnes’s table gives α_G ≈ 5.9×10⁻³⁹. arxiv +2 Gravity’s weakness is what allows large, long-lived stars and planets and big brains. Rees argues that a much smaller N would give only a “short-lived and miniature universe.” pratclif Caveat: Adams (2019) shows that stars can function over a considerable range of G and α, so the claims of “one part in 10^40 precision” circulating in popular apologetics are exaggerated. stanford arxiv

(f) The fine-structure constant (α ≈ 1/137) and the electron-to-proton mass ratio (β ≈ 1/1836). Barnes (2012) reproduces the classic α–β plane, where the life-permitting region is bounded by the following requirements: arxiv

stable atoms and ordered molecular structure (β must be small);
stars that are able to burn;
the existence of both convective and radiative stars;
nuclear stability.

Hawking makes the point in A Brief History of Time (1988, p. 125): “if the electric charge of the electron had been only slightly different, stars either would have been unable to burn hydrogen and helium, or else they would not have exploded.” aish homestead

(g) Hoyle’s carbon resonance and oxygen. In 1953 Hoyle predicted, and Fowler’s group confirmed, the excited state of carbon-12 that makes the triple-alpha process efficient. Huang, Adams & Grohs (2019) note that “The relevant excited state has an energy of 7.6444 MeV,” and that “the 8Be nucleus fails to have a bound state by only 92 keV.” Hoyle was struck by its relation to the 7.12 MeV level in oxygen-16: “If you wanted to produce carbon and oxygen in roughly equal quantities by stellar nucleosynthesis, these are just the two levels you would have to fix” (Engineering & Science, Nov. 1981, p. 12). caltech

Oberhummer, Csótó & Schlattl (Science 289: 88–90, 2000) found that “outside a narrow window of 0.5 and 4% of the values of the strong and Coulomb forces, respectively, the stellar production of carbon or oxygen is reduced by factors of 30 to 1000.” They also found that such changes “would destroy either nearly all C or all O in every star.” semanticscholar arxiv

Caveat (important): Weinberg (“A Designer Universe?”, NYRB, 1999) argued that the resonance is “not such a close call after all.” physlink Huang, Adams & Grohs (Astroparticle Physics 105, 2019: 13–24) ran ~2,400 stellar simulations of 15–40 solar-mass stars. They found significant carbon for “-300 keV ≲ ΔE_R ≲ 500 keV” and a narrower oxygen range of “-300 keV ≲ ΔE_R ≲ 300 keV.” By their estimate, the ~800 keV carbon window “corresponds to ∼10−20% changes to the parameters of fundamental physics.” This is a case where honest defenders should concede that the margin is larger than the 1980s literature implied.

(h) Deuteron binding and diproton stability. If the strong force were somewhat stronger, the diproton (²He) would be bound. That was classically said to burn all hydrogen in the Big Bang (Dyson 1971). Barnes himself notes that this “may not be as clear-cut a disaster as is often asserted.” MacDonald & Mullan (Phys. Rev. D 80, 043507, 2009) found that a 13% increase in the strong coupling binds the diproton, yet “provided the increase in strong force coupling constant is less than about 50% substantial amounts of hydrogen remain.” The deuteron, by contrast, is barely bound (~2.2 MeV). In Barr & Khan’s map, modest changes in quark masses make it unstable, which breaks the first step of stellar hydrogen burning. arxiv

(i) Expansion rate and flatness. Hawking (A Brief History of Time, 1988, ~p. 121): “If the rate of expansion one second after the big bang had been smaller by even one part in a hundred thousand million million, the universe would have recollapsed before it ever reached its present size.” kwize +2 Barnes’s calculation gives 1 part in ~10^16 at one second and ~10^60 at the Planck time. arxiv

Caveat: Hawking himself noted a few pages later that inflation could make the expansion rate “automatically become very close to the critical rate.” arxiv arxiv Carroll argued in 2014 that a proper measure on cosmologies dissolves this tuning. reasonablefaith But Carroll & Tam (2010) also found that the fraction of realistic cosmologies eligible for inflation is “P(inflation) ≈ 10^(−6.6×10^7),” so inflation relocates the tuning rather than abolishing it. arxiv Treat flatness as a demonstration that one tuning may have a dynamical explanation, not as a lead argument.

(j) Matter–antimatter asymmetry, dimensions, and the laws themselves. The early universe had roughly one extra baryon per billion baryon–antibaryon pairs. Sakharov’s conditions show that this requires specific symmetry violations, and Adams finds that the baryon-to-photon ratio can vary over a sizable range. researchgate arxiv On dimensionality, Rees notes that “Life couldn’t exist if D were two or four” pratclif (with three spatial dimensions, stable orbits and atoms are possible). Tegmark’s work on the dimensionality of spacetime extends this to time dimensions. Barnes also catalogs fine-tuning of the form of the laws. If electrons were bosons there would be no Pauli exclusion and hence no chemistry. If the strong force had long range there would be no atoms. And Conway had to search through many rule sets before the Game of Life “was effected.” arxiv

3. The Skeptical Objections and the Strongest Responses

Objection 1: “There is no fine-tuning; it is all explainable by known physics” (Victor Stenger, The Fallacy of Fine-Tuning, 2011). Stenger claimed that “the most commonly cited examples of apparent fine-tuning can be readily explained by the application of a little well-established physics and cosmology.” Barnes’s 2012 review is a line-by-line rebuttal. cambridge Its main charges:

Stenger conflates covariance with symmetry when he tries to derive the laws from “point-of-view invariance.”
His entropy argument “presented the problem itself as its solution,” because it simply assumes homogeneity.
His “wedge” argument, that varying parameters together opens up room, commits what Barnes calls the “sequential juggler” and “flippant funambulist” fallacies. Real parameter-space studies show “many intersecting wedges,” not one. arxiv

Barnes also lists scientists of all persuasions who accept fine-tuning: reasonablefaith “Barrow, Carr, Carter, Davies, Dawkins, Deutsch, Ellis, Greene, Guth, Harrison, Hawking, Linde, Page, Penrose, Polkinghorne, Rees, Sandage, Smolin, Susskind, Tegmark, Tipler, Vilenkin, Weinberg, Wheeler, Wilczek.” arxiv Stenger’s position is a minority one.

Objection 2: “Life could adapt to other conditions.” This is Carroll’s first point in 2014: “I will start granting that once someone tells me the conditions under which life can exist.” crossexaminedblog reasonablefaith The response is that the catastrophes in Tier 1 do not merely alter Earth-like biology. They eliminate any complexity:

no atoms other than hydrogen, or no stable atoms at all;
universes that recollapse in seconds or that never form bound structures;
only black holes.

Barnes: “however many ways there may be of being alive… there are vastly more ways of being dead” (quoting Dawkins). arxiv Any information-processing system needs stable, complex structures and long-lived energy sources.

Objection 3: Fred Adams’s “stars in other universes” argument. Adams (Physics Reports 807, 2019) concludes that “For all of these issues, viable universes exist over a range of parameter space, which is delineated herein. Finally, for universes with significantly different parameters, new types of astrophysical processes can generate energy and support habitability.” He also argues that our universe “could have been more, rather than less, life-friendly” (SEP summary). This is the most serious scientific moderation, and it should be conceded where it applies (stellar parameters, the carbon resonance, the baryon ratio). Adams does not touch the cosmological constant’s ~120-order problem or the entropy problem. He also acknowledges that some cancellations are “not automatic,” so that they require “some measure of fine-tuning.” sciencedirect His work shrinks the list of tunings; it does not empty it.

Objection 4: The normalizability problem (McGrew, McGrew & Vestrup, Mind 110 (2001): 1027–1038). The objection holds that if a constant could take any real value, a uniform probability over an infinite range cannot be normalized, so “the narrow intervals do not yield a probability at all.” philpapers There are three responses:

Robin Collins (in the Blackwell Companion to Natural Theology, 2009) restricts the argument to the “epistemically illuminated range” that physics actually lets us evaluate.
Physical theories supply natural scales. The Planck scale bounds masses and energies, so ranges are finite in practice.
Barnes and colleagues (arXiv:1707.03965; Barnes, “A Reasonable Little Question,” Ergo 6, 2019) show that the normalizability worry, if valid, would equally undercut ordinary Bayesian model selection throughout physics. arxiv philarchive Physicists in fact routinely call parameters “unnatural” on precisely this basis.

Objection 5: “The weak anthropic principle makes it unsurprising.” Barnes treats the anthropic principle as a genuine selection effect, like Malmquist bias in astronomy, but notes that “a selection bias alone cannot explain anything.” arxiv John Leslie’s firing-squad parable (Universes, 1989; also used by Francis Collins, The Language of God, p. 77) makes the point. Fifty marksmen fire and all miss. gsu You should not say, “Of course they missed, or I wouldn’t be here to notice.” You should ask whether the miss was arranged. Observing the outcome is guaranteed if you survive, but the outcome itself still needs explaining.

Objection 6: “Theism doesn’t predict fine-tuning; God could make life any way” (Carroll). Carroll argued that “God doesn’t need to fine-tune anything,” and that the early universe’s entropy is far lower than life needs. scienceandculture reasonablefaith Three replies:

Only an embodied, law-governed, discoverable world allows creatures who learn, act and are morally responsible. A designer with those aims has reason to work through precise, stable laws rather than ad hoc miracle.
Collins argues that over-tuned low entropy serves the discoverability of the cosmos. reasonablefaith
In Islamic terms, God’s sunna, His custom in creation (sunnat Allah), is regular and intelligible by design. (See the Epilogue.)

4. The Explanatory Options

Chance in a single universe. Few defend this today. Flew (with Varghese), There Is a God (2007, p. 115): “Virtually no major scientist today claims that the fine tuning was purely a result of chance factors at work in a single universe.” distantreader distantreader
Physical necessity (a deeper theory fixes the values). This is possible, but string theory went the opposite way, toward a “landscape” of ~10^500 vacua. arxiv reasonablefaith Barnes notes that a dynamical explanation like inflation works only where a physically special value happens to lie inside the life-permitting range; “there can be no inflation-like dynamical solution” for quark masses or force strengths. arxiv A necessitating deeper law would raise its own question: why is the necessary law life-permitting?
Multiverse plus anthropic selection. This is the main naturalistic answer (Susskind, Weinberg, Rees, Linde, Vilenkin, Tegmark, Carroll). Its burdens:
- The measure problem. In an infinite multiverse, probabilities are ill-defined without a contested cutoff. arxiv
- Boltzmann brains. Many multiverse models predict that random fluctuating observers vastly outnumber evolved ones. That would make our ordered observations improbable, a problem Carroll concedes must be engineered away model by model. reasonablefaith
- The inverse gambler’s fallacy. Ian Hacking (“The Inverse Gambler’s Fallacy,” Mind 96, 1987: 331–340) and Roger White (“Fine-Tuning and Multiple Universes,” Noûs 34, 2000: 260–276) argue that the existence of many other universes does not make it more probable that this one is life-permitting. cambridge John Leslie (“No Inverse Gambler’s Fallacy in Cosmology,” Mind 97, 1988) and others dispute this, and the debate is unresolved. wiley inspirehep
- Testability. George Ellis and Joe Silk (“Scientific method: Defend the integrity of physics,” Nature 516, 2014: 321–323) insist that “a theory must be falsifiable to be scientific” and warn against “abandoning testability for elegance.” researchgate uncommondescent
- The cane-toad problem. The universe-generating mechanism, such as inflation plus a string landscape, must itself be life-permitting. medium richardcarrier Barnes warns that the “solution” may be “more fine-tuned than the problem.” arxiv
Design. On this view the life-permitting values reflect intention. Its strength is that it predicts what the data show: a universe whose arbitrary parameters land in life-permitting and discoverable ranges. Its weakness, in the eyes of critics, is its lack of quantitative predictive specificity (Carroll’s “theism is not well defined”). reasonablefaith Note the asymmetry, however. The multiverse’s predictive power also depends on an undetermined measure.

Verdict. Chance is dead. Necessity is a promissory note. The real contest is between multiverse and design, and both are metaphysically weighty. Fine-tuning raises the probability of design relative to single-universe naturalism by an enormous factor. Whether it beats the multiverse depends on prior commitments that physics alone cannot settle.

5. What the Experts Actually Said: Verified Quotations

Quotes have been checked against the primary text or a reliable secondary source, as noted. Quotes that could not be verified are listed separately and should not be used.

(a) Theistic or design-leaning interpretations

Fred Hoyle (agnostic, not a conventional theist). “A common sense interpretation of the facts suggests that a superintellect has monkeyed with physics, as well as with chemistry and biology, and that there are no blind forces worth speaking about in nature. The numbers one calculates from the facts seem to me so overwhelming as to put this conclusion almost beyond question.” Engineering & Science (Caltech), November 1981, p. 12 (verified from the primary PDF; a similar passage appears in Annual Review of Astronomy and Astrophysics 20, 1982). gsu +2 Same article: “if you were a sensible superintellect you would conclude that the carbon atom is a fix.” caltech
Freeman Dyson. “The more I examine the universe and study the details of its architecture, the more evidence I find that the universe in some sense must have known that we were coming.” Disturbing the Universe (1979), p. 250. His careful qualifier follows: “I do not claim that the architecture of the universe proves the existence of God. I claim only that the architecture of the universe is consistent with the hypothesis that mind plays an essential role in its functioning.” goodreads scienceandculture (Reliable secondary. A widely circulated variant, “it almost seems as if the Universe must in some sense have known…”, is a different, later formulation; do not merge the two.) todayinsci +2
Arno Penzias (Nobel laureate, CMB co-discoverer). “In summary, therefore, astronomy leads us to a unique event, a universe which was created out of nothing, one with the very delicate balance needed to provide exactly the conditions required to permit life, and one which has an underlying (one might say ‘supernatural’) plan.” This comes from his 1983 Handleman Prize address, reprinted in Margenau & Varghese (eds.), Cosmos, Bios, Theos (1992), p. 83 (reliable secondary). gsu +2 Flag: the popular version containing “absurdly improbable accident” is a garbled paraphrase, and it is not from the 1978 New York Times. azquotes citatis
Antony Flew (former leading atheist philosopher; written with Roy Varghese, who reportedly drafted much of the text). “Virtually no major scientist today claims that the fine tuning was purely a result of chance factors at work in a single universe” (There Is a God, 2007, p. 115). He also writes: “it is a hard fact that we live in a universe with certain laws and constants, and life would not have been possible if some of these laws and constants had been different… There is currently no evidence in support of a multiverse. It remains a speculative idea” (p. 119). distantreader
Francis Collins. “for those willing to consider a theistic perspective, the Anthropic Principle certainly provides an interesting argument in favor of a Creator.” The Language of God (2006), p. 78 (reliable secondary). credohouse Flag: a “15 constants… knew we were coming” passage often attributed to this book comes via Tim Keller’s The Reason for God (2008) and is unverified as book text. A Goodreads line credited to Collins is Dyson’s. goodreads +2

(b) Naturalists who concede that fine-tuning is real and explain it otherwise

Stephen Hawking. “The remarkable fact is that the values of these numbers seem to have been very finely adjusted to make possible the development of life.” A Brief History of Time (1988), p. 125. medium +2 On the expansion rate (~p. 121): “smaller by even one part in a hundred thousand million million, the universe would have recollapsed.” (Context: he immediately weighs anthropic and inflationary explanations.) kwize arxiv
Steven Weinberg. “the existence of life of any kind seems to require a cancellation between different contributions to the vacuum energy, accurate to about 120 decimal places” (Scientific American, Oct. 1994). aish nideffer Weinberg was skeptical of most claimed tunings (“not impressed with these supposed instances of fine-tuning”) but made one exception: the cosmological constant “does seem remarkably well adjusted in our favor” (“A Designer Universe?”, NYRB, 1999). beliefmap physlink His naturalistic answer: “the string landscape may explain how the constants of nature that we observe can take values suitable for life without being fine-tuned by a benevolent creator” (“Living in the Multiverse,” 2005; in Carr (ed.), Universe or Multiverse?, 2007). arxiv arxiv
Leonard Susskind. “To make the first 119 decimal places of the vacuum energy zero is most certainly no accident.” The Cosmic Landscape: String Theory and the Illusion of Intelligent Design (2005). The subtitle states his anti-design conclusion. Wikipedia +2
Martin Rees (Astronomer Royal). “These six numbers constitute a ‘recipe’ for a universe.” He continues: “Is this tuning just a brute fact, a coincidence?… An infinity of other universes may well exist where the numbers are different.” firstscience Just Six Numbers (1999), ch. 1. He favors a multiverse. focus
Paul Davies. “Scientists are slowly waking up to an inconvenient truth – the universe looks suspiciously like a fix. goodreads The issue concerns the very laws of nature themselves.” The headline of the same piece states his position: “Yes, the universe looks like a fix. But that doesn’t mean that a god fixed it” (The Guardian, 25 June 2007). azquotes Flag: “The seemingly miraculous concurrence of these numerical values must remain the most compelling evidence for cosmic design” is widely attributed to Davies azquotes but was not verified to a page here; use with caution. forumterrace
Roger Penrose. “This now tells us how precise the Creator’s aim must have been: namely to an accuracy of one part in 10^10^123.” The Emperor’s New Mind (1989), ch. 7 (pp. ~339–345). ws5 The “Creator” is figurative.
Max Tegmark, Anthony Aguirre, Martin Rees, Frank Wilczek. “We identify 31 dimensionless physical constants required by particle physics and cosmology, and emphasize that both microphysical constraints and selection effects might help elucidate their origin” (Phys. Rev. D 73, 023505, 2006). elsevierpure This is a Nobel laureate (Wilczek) and an MIT cosmologist treating anthropic selection as a working scientific tool.
Luke Barnes (Christian, but writing as a technical reviewer). “We conclude that the universe is fine-tuned for the existence of life. Of all the ways that the laws of nature, constants of physics and initial conditions of the universe could have been, only a very small subset permits the existence of intelligent life” (PASA 2012, §6). arxiv

(c) Skeptics who concede that it is a serious argument

Sean Carroll. “I’m very happy to admit right off the bat – this is the best argument that the theists have when it comes to cosmology… Given that, it is still a terrible argument.” On entropy: “I’m not saying there’s not fine-tunings; I’m saying they’re not there for life.” (Debate with W. L. Craig, Greer-Heard Forum, New Orleans, March 2014; verified from the published transcript.) crossexaminedblog reasonablefaith In From Eternity to Here (2010), as quoted by Craig: “If the universe we see is really all there is with the Big Bang as a low-entropy beginning, we seem to be stuck with an uncomfortable fine-tuning problem.” reasonablefaith
Christopher Hitchens. “I think every one of us picks the fine-tuning one as the most intriguing… It’s not a trivial [argument].” From the documentary Collision (2009), transcribed from film footage. The bracketed word is the transcriber’s, and the wording is only reliable at the secondary level. Hitchens added in the same breath that “it doesn’t prove design, doesn’t prove a Designer.” quora wordpress
Richard Dawkins. “Physicists have calculated that if the laws and constants of physics had been even slightly different, the universe would have developed in such a way that life would have been impossible.” The God Delusion (2006), ch. 4, p. 141 (UK ed.). focus His answer: “Some kind of multiverse theory could in principle do for physics the same explanatory work as Darwinism does for biology.” Wikipedia
Victor Stenger is the clearest example of an outright denier (“the most commonly cited examples of apparent fine-tuning can be readily explained…”). His position is rebutted in detail by Barnes (2012). arxiv cambridge

Not quote-verified in this review (use paraphrase only)

John Wheeler (participatory universe), Brandon Carter, Bernard Carr, Andrei Linde and Alexander Vilenkin (eternal inflation as the explanation), Lee Smolin (cosmological natural selection), George Ellis, Geraint Lewis, Robin Collins, John Leslie, Richard Swinburne, Thomas Nagel (Mind and Cosmos, 2012), Daniel Dennett, Charles Townes, John Polkinghorne, Owen Gingerich and Allan Sandage all acknowledged or engaged fine-tuning, and many appear on Barnes’s list of scientists who accept it. No exact wording with page numbers for them was verified here, so they should be cited by stance, not by quotation, until checked against the books themselves.

Recommendations for the Conversation with a Skeptic

Separate the datum from the inference. Ask the skeptic whether they deny that the life-permitting range is small (a physics question) or deny that this points to God (a philosophy question). Most well-read skeptics, following Carroll, Weinberg and Dawkins, concede the first.
Lead with the cosmological constant and initial entropy. These are the hardest to dissolve and are conceded by atheist physicists. Avoid opening with the carbon resonance or flatness.
Quote the naturalists, not the apologists. Weinberg, Susskind, Carroll, Rees and Hawking carry more weight with skeptics than Craig or Strobel.
Concede Adams’s points gracefully. Doing so strengthens credibility and isolates the robust core.
Press the multiverse on its own costs: the measure problem, Boltzmann brains, the inverse gambler’s fallacy, testability, and the fine-tuning of the generator itself.
Claim no more than the evidence supports. The defensible claim is that fine-tuning makes design significantly more probable than single-universe chance. The claim that it proves God goes beyond the evidence.